AB 642: Law enforcement agencies: facial recognition technology.
- Session Year: 2023-2024
- House: Assembly
- Latest Version Date: 2023-05-01
Current Status:
Failed
(2024-02-01: From committee: Filed with the Chief Clerk pursuant to Joint Rule 56.)
Introduced
In Committee
First Chamber
In Committee
Second Chamber
Enacted
Existing law, generally, regulates state and local law enforcement agencies regarding subject matter that includes the selection and training of peace officers, the maintenance and release of records, the use of force, and the use of certain equipment. Previous law, until January 1, 2023, prohibited the use of real-time facial recognition technology by law enforcement agencies in connection with body-worn cameras.
This bill would prescribe the acceptable and prohibited uses for facial recognition technology (FRT), as defined, by a law enforcement agency or peace officer, as defined. The bill would set certain requirements for FRT systems and reference databases, as defined, used by law enforcement agencies. The bill would require the Department of Technology, in consultation with the Chief of the Office of Information Security, to issue standards to ensure the confidentiality and cybersecurity of FRT data and results. The bill would require law enforcement agencies to establish written policies that adhere to those standards. The bill would require law enforcement agencies utilizing FRT to keep certain records and to annually report certain data to the California State Auditor. The bill would require any law enforcement agency utilizing FRT to have a written policy, as specified, and to post that policy on their internet website. The bill would require a law enforcement agency to provide certain information to any person that FRT is used to identify.
The bill would also require each district attorneys office, city prosecutors office, and the Attorney General to annually report certain data to the California State Auditor.
By imposing a new reporting requirement on district attorneys and city prosecutors, this bill would impose a state-mandated local program.
The bill would require the California State Auditor to annually analyze the data submitted and to prepare and publicize a report, as specified.
The bill would authorize a person harmed by a violation of these provisions to bring a civil action for damages, including statutory damages, as specified.
The California Constitution requires the state to reimburse local agencies and school districts for certain costs mandated by the state. Statutory provisions establish procedures for making that reimbursement.
This bill would provide that, if the Commission on State Mandates determines that the bill contains costs mandated by the state, reimbursement for those costs shall be made pursuant to the statutory provisions noted above.
Existing law, generally, regulates state and local law enforcement agencies regarding subject matter that includes the selection and training of peace officers, the maintenance and release of records, the use of force, and the use of certain equipment. Previous law, until January 1, 2023, prohibited the use of real-time facial recognition technology by law enforcement agencies in connection with body-worn cameras.
This bill would prescribe the acceptable and prohibited uses for facial recognition technology (FRT), as defined, by a law enforcement agency or peace officer, as defined. The bill would set certain requirements for FRT systems and reference databases, as defined, used by law enforcement agencies. The bill would require law enforcement agencies utilizing FRT to keep certain records and to annually report certain data to the California State Auditor. The bill would require any law enforcement agency utilizing FRT to have a written policy, as specified, and to post that policy on their internet website. The bill would require a law enforcement agency to provide certain information to any person that FRT is used to identify.
The bill would also require each district attorneys office, city prosecutors office, and the Attorney General to annually report certain data to the California State Auditor.
By imposing a new reporting requirement on district attorneys and city prosecutors, this bill would impose a state-mandated local program.
The bill would require the California State Auditor to annually analyze the data submitted and to prepare and publicize a report, as specified.
The bill would authorize a person harmed by a violation of these provisions to bring a civil action for damages, including statutory damages, as specified.
The California Constitution requires the state to reimburse local agencies and school districts for certain costs mandated by the state. Statutory provisions establish procedures for making that reimbursement.
This bill would provide that, if the Commission on State Mandates determines that the bill contains costs mandated by the state, reimbursement for those costs shall be made pursuant to the statutory provisions noted above.
Existing law, generally, regulates state and local law enforcement agencies regarding subject matter that includes the selection and training of peace officers, the maintenance and release of records, the use of force, and the use of certain equipment. Previous law, until January 1, 2023, prohibited the use of real-time facial recognition technology by law enforcement agencies in connection with body-worn cameras.
This bill, commencing July 1, 2024, would require any law enforcement agency, as defined, that uses facial recognition technology (FRT), as defined, to have a written policy governing the use of that technology. The bill would require any FRT system used to meet certain national standards and would limit the use of FRT to use as an investigative aid, as described. The bill would specifically prohibit the use of any FRT-generated match from being the sole basis for probable cause in an arrest, search, or warrant. The bill would also require an agency using FRT to post their written policy and an annual summary of FRT usage, as specified, on their internet website.This bill would prescribe the acceptable and prohibited uses for facial recognition technology (FRT), as defined, by a law enforcement agency or peace officer, as defined. The bill would set certain requirements for FRT systems and reference databases, as defined, used by law enforcement agencies. The bill would require law enforcement agencies utilizing FRT to keep certain records and to annually report certain data to the California State Auditor. The bill would require any law enforcement agency utilizing FRT to have a written policy, as specified, and to post that policy on their internet website. The bill would require a law enforcement agency to provide certain information to any person that FRT is used to identify.
The bill would also require each district attorneys office, city prosecutors office, and the Attorney General to annually report certain data to the California State Auditor.
By imposing a new reporting requirement on district attorneys and city prosecutors, this bill would impose a state-mandated local program.
The bill would require the California State Auditor to annually analyze the data submitted and to prepare and publicize a report, as specified.
The bill would authorize a person harmed by a violation of these provisions to bring a civil action for damages, including statutory damages, as specified.
The California Constitution requires the state to reimburse local agencies and school districts for certain costs mandated by the state. Statutory provisions establish procedures for making that reimbursement.
This bill would provide that, if the Commission on State Mandates determines that the bill contains costs mandated by the state, reimbursement for those costs shall be made pursuant to the statutory provisions noted above.
Existing law, generally, regulates state and local law enforcement agencies regarding subject matter that includes the selection and training of peace officers, the maintenance and release of records, the use of force, and the use of certain equipment. Previous law, until January 1, 2023, prohibited the use of real-time facial recognition technology by law enforcement agencies in connection with body-worn cameras.
This bill, commencing July 1, 2024, would require any law enforcement agency, as defined, that uses facial recognition technology (FRT), as defined, to have a written policy governing the use of that technology. The bill would require any FRT system used to meet certain national standards and would limit the use of FRT to use as an investigative aid, as described. The bill would specifically prohibit the use of any FRT-generated match from being the sole basis for probable cause in an arrest, search, or warrant. The bill would also require an agency using FRT to post their written policy and an annual summary of FRT usage, as specified, on their internet website.
Existing law, generally, regulates state and local law enforcement agencies regarding subject matter that includes the selection and training of peace officers, the maintenance and release of records, the use of force, and the use of certain equipment. Previous law, until January 1, 2023, prohibited the use of real-time facial recognition technology by law enforcement agencies in connection with body-worn cameras.This bill would express the intent of the Legislature to enact legislation relating to the use of facial recognition technology by law enforcement agencies.Existing law, generally, regulates state and local law enforcement agencies regarding subject matter that includes the selection and training of peace officers, the maintenance and release of records, the use of force, and the use of certain equipment. Previous law, until January 1, 2023, prohibited the use of real-time facial recognition technology by law enforcement agencies in connection with body-worn cameras.
This bill would express the intent of the Legislature to enact legislation relating to the use of facial recognition technology by law enforcement agencies.
As used in this title, the following terms are defined as follows:
(a)Arrest photograph database means a database populated primarily by booking or arrest photographs or other photographs of persons with law enforcement contacts.
(b)(1)Facial recognition technology or FRT means a system that compares a probe image of an unidentified human face against a reference photograph database, and, based on biometric data, generates possible matches to aid in identifying the person in the probe image.
(2)FRT includes any surveillance system that actively uses FRT to identify persons in a surveilled area in real time.
(3)FRT does not include any access control system used by a law enforcement agency that uses biometric inputs to confirm the identity of employees or other approved persons for the purpose of controlling access to any secured place, device, or system.
(c)Law enforcement agency means any department or agency of the state or any political subdivision thereof that employs any peace officer.
(d)Peace officer means any officer described in Section 830.
(e)Probe image means an image of a person that is searched against a database of known, identified persons or an unsolved photograph file.
(f)Reference photograph database means a database populated with photographs of individuals that have been identified, including databases composed of drivers licenses or other documents made or issued by or under the authority of the state, a political subdivision thereof, databases operated by third parties, and arrest photograph databases.
13675.1.(a)A peace officer is authorized to use or request the use of FRT for any of the following reasons:
(1)To assist in identifying a person that the officer has reasonable suspicion to believe has committed a felony.
(2)To assist in identifying any person who is deceased or who has been reported missing pursuant to Section 14211.
(3)To assist in identifying any person who has been lawfully arrested, during the process of booking or during that persons custodial detention.
(4)To assist in identifying any person if a peace officer determines that an emergency situation exists that involves immediate danger of death or serious physical injury to any person and the identification of that person is necessary to prevent that death or injury.
(b)If a peace officer uses or requests the use of FRT pursuant to subdivision (a), the officer shall document all of the following information:
(1)The identity of the peace officer using or requesting the use of FRT, and, if applicable, the officer authorizing the use or request.
(2)A detailed description, as available, of the person being identified.
(3)Any photograph or video being used as the probe image.
(4)Any details regarding other investigative measures taken to identify the person and an explanation of why those measures failed or are reasonably unlikely to succeed.
(c)(1)Beginning on July 1, 2024, and every six months thereafter, the custodian of any arrest photograph database being used in conjunction with FRT shall remove from the database any of the following images:
(A)Any photograph of a person under 18 years of age.
(B)Any photograph of an arrested or detained person who has been released without being charged with an offense, or whose charges have been dismissed.
(C)Any photograph of an arrested or detained person who has subsequently been acquitted of the charged offense.
(D)Any photograph of a person whose conviction has been expunged, has been exonerated of the crime, or who has had their conviction reversed on appeal.
(2)This section applies only to the use of a reference photograph database for the use of FRT and shall not be construed to prohibit a peace officer from using any other investigative database, including a fingerprint database.
(3)Any agency that maintains and operates an arrest photograph database shall establish procedures to ensure compliance with this subdivision.
13675.2.A peace officer using or requesting the use of A law enforcement agency or officer authorizing the use of, maintaining the database for, or using FRT shall not do any of the following:
(a)Use FRT to identify any person solely on the basis that the person is exercising rights guaranteed by the United States Constitution, including free assembly, association, and speech.
(b)Rely on actual or perceived race, ethnicity, national origin, religion, disability, gender, gender identity, or sexual orientation in selecting a person to identify using FRT, except when there is trustworthy information, relevant to the locality and timeframe, in the context of a particular area and for a particular period of time, that links a person with a particular characteristic described in this subdivision to an identified criminal incident or scheme.
(c)Share FRT data with any state or federal agency for the purpose of enforcing federal immigration law.(d)(1)Provide the results of, or information derived from, the use of FRT to any individual or to any agency or department in another state regarding the provision of lawful gender-affirming health care or gender-affirming mental health care performed in this state.(c)Notwithstanding any other provision of this title, release any FRT data or results, or allow access to the databases used with FRT, in response to a subpoena or request if that subpoena or request is in violation of the California Values Act (Chapter 17.25 (commencing with Section 7284) of Division 7 of Title 1 of the Government Code).
(d)(1)Notwithstanding any other provision of this title, release any FRT data or results, or allow access to databases used with FRT, in response to a subpoena or request if that subpoena or request is based on any of the following:
(A)Another states laws that interfere with a persons rights under the Reproductive Privacy Act (Article 2.5 (commencing with Section 123460) of Chapter 2 of Part 2 of Division 106 of the Health and Safety Code).
(B)A foreign penal civil action, as defined in Section 2029.200 of the Code of Civil Procedure.
(C)An investigation related to a natural person seeking gender-affirming health care or gender-affirming mental health care.
(2)As used in this subdivision, gender affirming health care and gender affirming mental health care gender-affirming health care and gender-affirming mental health care have the same meaning as in Section 16010.2 of the Welfare and Institutions Code.
(e)(1)Provide the results of, or information derived from, the use of FRT to any individual or agency or department in another state regarding the provision of abortion services in this state.
(2)For purposes of this subdivision and subdivision (d), information derived from the use of FRT means information that would not have been discovered or obtained but for the use of FRT, regardless of any claim that the information would inevitably have been discovered or obtained through other means.
(f)Use an FRT match as the sole basis upon which probable cause is established for a search, arrest, or affidavit for a warrant. Any peace officers using information obtained from the use of FRT shall examine results with care and consider the possibility that matches could be inaccurate.
(g)Use FRT in conjunction with any reference photograph database that contains information, including images, obtained by any of the following means:
(1)In a manner that violates federal or state law.
(2)In a manner that violates a service agreement between a provider of an electronic communication service to the public or a provider of a remote computing service and customers or subscribers of that provider.
(3)In a manner that is inconsistent with the privacy policy of a provider described in paragraph (2), if applicable.
(4)By deceiving a person whose information was obtained.
(5)Through the unauthorized access of an electronic device or online account.
(6)In violation of a contract, court settlement, or other binding legal agreement.
(7)From unlawful or unconstitutional practices by any governmental official or entity.
13675.3.(a)By no later than January 31, 2025, and annually thereafter, each law enforcement agency using FRT, including agencies whose officers use or request the use of the FRT of another agency, shall prepare and submit a report to the California State Auditor containing only the following information regarding the use of FRT, as applicable:
(1)The information required to be documented by subdivision (b) of Section 13675.1 for each use or request for facial recognition.
(2)Whether modifications were made to any probe images and what those modifications were.
(3)The arrests that FRT results contributed to, and the offenses for which the arrests were made, disaggregated by race, ethnicity, gender, and age.
(4)A description of the reference photograph database that was used.
(5)A description of the FRT system that was used.
(6)The total number of searches performed.
(7)The total number of times FRT results were shared with another law enforcement agency, an outside agency, or a third party.
(8)Actions taken to comply with subdivision (c) of Section 13675.1 regarding unusable arrest photographs.
(b)By no later than January 31, 2025, and annually thereafter, each district attorneys office, city prosecutors office, and the Attorney General shall report to the California State Auditor the following information regarding information obtained from FRT:
(1)The number of convictions that FRT results contributed to and the offenses for which the convictions were obtained, disaggregated by race, ethnicity, gender, and age.
(2)The number of motions to suppress made related to FRT results, and the number granted or denied.
(c)On or before July 1 of each year, the California State Auditor shall release to the public, post online, and transmit to the Legislature a full and complete report concerning the use of FRT by law enforcement agencies and prosecutors, including any violations of this title that were identified.
(d)Any law enforcement agency using FRT shall keep and maintain those activity logs or other records necessary to comply with this section.
13675.4.(a)(1)A law enforcement agency shall not operate an FRT system that has not been evaluated under the National Institute of Standards and Technology Face Recognition Vendor Testing Program and achieved an accuracy score of 98 percent true positives within two or more datasets relevant to investigative applications on a program report.
(2)A peace officer shall not use FRT in any manner that reduces the facial recognition programs competency score below 98 percent.
(b)(1)By July 1, 2024, the Department of Technology, in consultation with the Chief of the Office of Information Security, shall issue standards to ensure the confidentiality and cybersecurity of FRT data and results.
(b)(2)A law enforcement agency that uses FRT shall have a written policy that includes, adheres to the standards in paragraph (1). The policy shall include, without limitation, all of the following:
(1)(A)A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT by the Commission on Peace Officer Standards and Training.
(2)(B)A requirement that a manager authorized to use FRT be assigned to oversee the FRT program.
(3)(C)A policy that describes the parameters of acceptable inputs to be used as probe images and that prohibits the use of sketches or other manually produced images.
(4)(D)An acceptable use policy that includes specific allowances and restrictions on use.
(3)This subdivision does not prevent a law enforcement agency from adopting a written policy that provides for standards stronger than those issued by the department.
(c)By no later than July 1, 2024, each law enforcement agency using FRT shall post the written policy required by subdivision (b) on its internet website.
(d)No law enforcement agency or peace officer shall request or enter into an agreement with another law enforcement agency or other third party to perform FRT searches on behalf of the requesting officer or agency if the program operated by the other agency or party does not meet the requirements of this section.
(e)(1)If the California State Auditor, pursuant to Section 13675.3, identifies any violations of this title by a law enforcement agency, that law enforcement agency shall cease using FRT until all violations have been corrected.
(2)If the use of FRT is suspended pursuant to this subdivision, the law enforcement agency shall notify the public of the suspension.
13675.5.(a)A law enforcement agency that uses FRT to attempt to identify an individual who is arrested shall provide to the individual both of the following:
(1)A notice of the name of the law enforcement agency that operated the FRT system used, and the name of the database, if any, that was used to identify the individual.
(2)A copy of the accuracy or bias report required by subdivision (a) of Section 13675.4, each probe image that was used by the agency, any modifications made to the probe image, the candidate list, in rank order, produced by the facial recognition system, and any other documentation related to the use of FRT in the law investigation.
(b)The information required under subdivision (a) shall be provided to the person in an appropriate language for the person if they are not fluent or literate in English.
(c)As used in this section candidate list means the top images that a facial recognition system determines to most closely match a probe image.
13675.6.(a)If a court or law enforcement agency determines that a peace officer has violated any provision of this title, and the court or agency finds that the circumstances surrounding the violation raise serious questions about whether or not the officer acted intentionally with respect to the violation, the agency shall promptly initiate a proceeding to determine whether disciplinary action against the officer is warranted.
(b)Notwithstanding any other law, a violation of this title is not punishable as a crime.
(c)A person who is subject to identification or attempted identification through FRT in violation of this title may bring a civil action against the peace officer or law enforcement agency responsible for the violation.
(d)A person who is the subject of disparate treatment or adverse impact on the basis of race, ethnicity, gender, or age, whether individually or as a member of a class of individuals, due to use of FRT or any technological element, criteria, method, or design feature thereof, by a law enforcement agency, may bring a cause of action against the peace officer, law enforcement agency, or maker of the facial recognition or face surveillance technology responsible for the violation.
(e)The following relief may be recovered or obtained in an action under subdivision (c) or (d):
(1)Any actual damages sustained by that person as a result of the violation.
(2)The greater of either of the following:
(A)Statutory damages of fifty thousand dollars ($50,000) per violation.
(B)Profits earned as a result of each violation.
(3)Exemplary damages pursuant to Section 3294 of the Civil Code.
(4)Injunctive, equitable, or declaratory relief as may be appropriate, including preliminary injunctive relief.
(5)Costs of the action, together with reasonable attorneys fees.
(f)A civil action under subdivision (b) or (c) may not be commenced later than two years after the date upon which the claimant discovered or first had a reasonable opportunity to discover the violation.
SEC. 2.If the Commission on State Mandates determines that this act contains costs mandated by the state, reimbursement to local agencies and school districts for those costs shall be made pursuant to Part 7 (commencing with Section 17500) of Division 4 of Title 2 of the Government Code.
As used in this title, the following terms are defined as follows:
(a)Arrest photograph database means a database populated primarily by booking or arrest photographs or other photographs of persons with law enforcement contacts.
(b)(1)Facial recognition technology or FRT means a system that compares a probe image of an unidentified human face against a reference photograph database, and, based on biometric data, generates possible matches to aid in identifying the person in the probe image.
(2)FRT includes any surveillance system that actively uses FRT to identify persons in a surveilled area in real time.
(3)FRT does not include any access control system used by a law enforcement agency that uses biometric inputs to confirm the identity of employees or other approved persons for the purpose of controlling access to any secured place, device, or system.
(c)Law enforcement agency means any department or agency of the state or any political subdivision thereof that employs any peace officer.
(d)Peace officer means any officer described in Section 830.
(e)Probe image means an image of a person that is searched against a database of known, identified persons or an unsolved photograph file.
(f)Reference photograph database means a database populated with photographs of individuals that have been identified, including databases composed of drivers licenses or other documents made or issued by or under the authority of the state, a political subdivision thereof, databases operated by third parties, and arrest photograph databases.
13675.1.(a)A peace officer is authorized to use or request the use of FRT for any of the following reasons:
(1)To assist in identifying a person that the officer has reasonable suspicion to believe has committed a felony.
(2)To assist in identifying any person who is deceased or who has been reported missing pursuant to Section 14211.
(3)To assist in identifying any person who has been lawfully arrested, during the process of booking or during that persons custodial detention.
(4)To assist in identifying any person if a peace officer determines that an emergency situation exists that involves immediate danger of death or serious physical injury to any person and the identification of that person is necessary to prevent that death or injury.
(b)If a peace officer uses or requests the use of FRT pursuant to subdivision (a), the officer shall document all of the following information:
(1)The identity of the peace officer using or requesting the use of FRT, and, if applicable, the officer authorizing the use or request.
(2)A detailed description, as available, of the person being identified.
(3)Any photograph or video being used as the probe image.
(4)Any details regarding other investigative measures taken to identify the person and an explanation of why those measures failed or are reasonably unlikely to succeed.
(c)(1)Beginning on July 1, 2024, and every six months thereafter, the custodian of any arrest photograph database being used in conjunction with FRT shall remove from the database any of the following images:
(A)Any photograph of a person under 18 years of age.
(B)Any photograph of an arrested or detained person who has been released without being charged with an offense, or whose charges have been dismissed.
(C)Any photograph of an arrested or detained person who has subsequently been acquitted of the charged offense.
(D)Any photograph of a person whose conviction has been expunged, has been exonerated of the crime, or who has had their conviction reversed on appeal.
(2)This section applies only to the use of a reference photograph database for the use of FRT and shall not be construed to prohibit a peace officer from using any other investigative database, including a fingerprint database.
(3)Any agency that maintains and operates an arrest photograph database shall establish procedures to ensure compliance with this subdivision.
13675.2.A peace officer using or requesting the use of FRT shall not do any of the following:
(a)Use FRT to identify any person solely on the basis that the person is exercising rights guaranteed by the United States Constitution, including free assembly, association, and speech.
(b)Rely on actual or perceived race, ethnicity, national origin, religion, disability, gender, gender identity, or sexual orientation in selecting a person to identify using FRT, except when there is trustworthy information, relevant to the locality and timeframe, in the context of a particular area and for a particular period of time, that links a person with a particular characteristic described in this subdivision to an identified criminal incident or scheme.
(c)Share FRT data with any state or federal agency for the purpose of enforcing federal immigration law.
(d)(1)Provide the results of, or information derived from, the use of FRT to any individual or to any agency or department in another state regarding the provision of lawful gender-affirming health care or gender-affirming mental health care performed in this state.
(2)As used in this subdivision, gender affirming health care and gender affirming mental health care have the same meaning as in Section 16010.2 of the Welfare and Institutions Code.
(e)(1)Provide the results of, or information derived from, the use of FRT to any individual or agency or department in another state regarding the provision of abortion services in this state.
(2)For purposes of this subdivision and subdivision (d), information derived from the use of FRT means information that would not have been discovered or obtained but for the use of FRT, regardless of any claim that the information would inevitably have been discovered or obtained through other means.
(f)Use an FRT match as the sole basis upon which probable cause is established for a search, arrest, or affidavit for a warrant. Any peace officers using information obtained from the use of FRT shall examine results with care and consider the possibility that matches could be inaccurate.
(g)Use FRT in conjunction with any reference photograph database that contains information, including images, obtained by any of the following means:
(1)In a manner that violates federal or state law.
(2)In a manner that violates a service agreement between a provider of an electronic communication service to the public or a provider of a remote computing service and customers or subscribers of that provider.
(3)In a manner that is inconsistent with the privacy policy of a provider described in paragraph (2), if applicable.
(4)By deceiving a person whose information was obtained.
(5)Through the unauthorized access of an electronic device or online account.
(6)In violation of a contract, court settlement, or other binding legal agreement.
(7)From unlawful or unconstitutional practices by any governmental official or entity.
13675.3.(a)By no later than January 31, 2025, and annually thereafter, each law enforcement agency using FRT, including agencies whose officers use or request the use of the FRT of another agency, shall prepare and submit a report to the California State Auditor containing only the following information regarding the use of FRT, as applicable:
(1)The information required to be documented by subdivision (b) of Section 13675.1 for each use or request for facial recognition.
(2)Whether modifications were made to any probe images and what those modifications were.
(3)The arrests that FRT results contributed to, and the offenses for which the arrests were made, disaggregated by race, ethnicity, gender, and age.
(4)A description of the reference photograph database that was used.
(5)A description of the FRT system that was used.
(6)The total number of searches performed.
(7)The total number of times FRT results were shared with another law enforcement agency, an outside agency, or a third party.
(8)Actions taken to comply with subdivision (c) of Section 13675.1 regarding unusable arrest photographs.
(b)By no later than January 31, 2025, and annually thereafter, each district attorneys office, city prosecutors office, and the Attorney General shall report to the California State Auditor the following information regarding information obtained from FRT:
(1)The number of convictions that FRT results contributed to and the offenses for which the convictions were obtained, disaggregated by race, ethnicity, gender, and age.
(2)The number of motions to suppress made related to FRT results, and the number granted or denied.
(c)On or before July 1 of each year, the California State Auditor shall release to the public, post online, and transmit to the Legislature a full and complete report concerning the use of FRT by law enforcement agencies and prosecutors, including any violations of this title that were identified.
(d)Any law enforcement agency using FRT shall keep and maintain those activity logs or other records necessary to comply with this section.
13675.4.(a)(1)A law enforcement agency shall not operate an FRT system that has not been evaluated under the National Institute of Standards and Technology Face Recognition Vendor Testing Program and achieved an accuracy score of 98 percent true positives within two or more datasets relevant to investigative applications on a program report.
(2)A peace officer shall not use FRT in any manner that reduces the facial recognition programs competency score below 98 percent.
(b)A law enforcement agency that uses FRT shall have a written policy that includes, without limitation, all of the following:
(1)A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT by the Commission on Peace Officer Standards and Training.
(2)A requirement that a manager authorized to use FRT be assigned to oversee the FRT program.
(3)A policy that describes the parameters of acceptable inputs to be used as probe images and that prohibits the use of sketches or other manually produced images.
(4)An acceptable use policy that includes specific allowances and restrictions on use.
(c)By no later than July 1, 2024, each law enforcement agency using FRT shall post the written policy required by subdivision (b) on its internet website.
(d)No law enforcement agency or peace officer shall request or enter into an agreement with another law enforcement agency or other third party to perform FRT searches on behalf of the requesting officer or agency if the program operated by the other agency or party does not meet the requirements of this section.
(e)(1)If the California State Auditor, pursuant to Section 13675.3, identifies any violations of this title by a law enforcement agency, that law enforcement agency shall cease using FRT until all violations have been corrected.
(2)If the use of FRT is suspended pursuant to this subdivision, the law enforcement agency shall notify the public of the suspension.
13675.5.(a)A law enforcement agency that uses FRT to attempt to identify an individual who is arrested shall provide to the individual both of the following:
(1)A notice of the name of the law enforcement agency that operated the FRT system used, and the name of the database, if any, that was used to identify the individual.
(2)A copy of the accuracy or bias report required by subdivision (a) of Section 13675.4, each probe image that was used by the agency, any modifications made to the probe image, the candidate list, in rank order, produced by the facial recognition system, and any other documentation related to the use of FRT in the law investigation.
(b)The information required under subdivision (a) shall be provided to the person in an appropriate language for the person if they are not fluent or literate in English.
(c)As used in this section candidate list means the top images that a facial recognition system determines to most closely match a probe image.
13675.6.(a)If a court or law enforcement agency determines that a peace officer has violated any provision of this title, and the court or agency finds that the circumstances surrounding the violation raise serious questions about whether or not the officer acted intentionally with respect to the violation, the agency shall promptly initiate a proceeding to determine whether disciplinary action against the officer is warranted.
(b)Notwithstanding any other law, a violation of this title is not punishable as a crime.
(c)A person who is subject to identification or attempted identification through FRT in violation of this title may bring a civil action against the peace officer or law enforcement agency responsible for the violation.
(d)A person who is the subject of disparate treatment or adverse impact on the basis of race, ethnicity, gender, or age, whether individually or as a member of a class of individuals, due to use of FRT or any technological element, criteria, method, or design feature thereof, by a law enforcement agency, may bring a cause of action against the peace officer, law enforcement agency, or maker of the facial recognition or face surveillance technology responsible for the violation.
(e)The following relief may be recovered or obtained in an action under subdivision (c) or (d):
(1)Any actual damages sustained by that person as a result of the violation.
(2)The greater of either of the following:
(A)Statutory damages of fifty thousand dollars ($50,000) per violation.
(B)Profits earned as a result of each violation.
(3)Exemplary damages pursuant to Section 3294 of the Civil Code.
(4)Injunctive, equitable, or declaratory relief as may be appropriate, including preliminary injunctive relief.
(5)Costs of the action, together with reasonable attorneys fees.
(f)A good faith reliance on a statutory authorization, or a good faith determination that the conduct complained of is permitted under this title, is a complete defense against any civil action brought under subdivision (c) or (d).(g)(f)A civil action under subdivision (b) or (c) may not be commenced later than two years after the date upon which the claimant discovered or first had a reasonable opportunity to discover the violation.
SEC. 2.If the Commission on State Mandates determines that this act contains costs mandated by the state, reimbursement to local agencies and school districts for those costs shall be made pursuant to Part 7 (commencing with Section 17500) of Division 4 of Title 2 of the Government Code.
As used in this title, the following terms are defined as follows:
(a)Arrest photograph database means a database populated primarily by booking or arrest photographs or other photographs of persons with law enforcement contacts.
(b)(1)Facial recognition technology or FRT means a system that compares a probe image of an unidentified human face against a reference photograph database, and, based on biometric data, generates possible matches to aid in identifying the person in the probe image.
(2)FRT includes any surveillance system that actively uses FRT to identify persons in a surveilled area in real time.
(3)FRT does not include any access control system used by a law enforcement agency that uses biometric inputs to confirm the identity of employees or other approved persons for the purpose of controlling access to any secured place, device, or system.
(c)Law enforcement agency means any department or agency of the state or any political subdivision thereof that employs any peace officer.
(d)Peace officer means any officer described in Section 830.
(e)Probe image means an image of a person that is searched against a database of known, identified persons or an unsolved photograph file.
(f)Reference photograph database means a database populated with photographs of individuals that have been identified, including databases composed of drivers licenses or other documents made or issued by or under the authority of the state, a political subdivision thereof, databases operated by third parties, and arrest photograph databases.
13675.1.(a)A peace officer is authorized to use or request the use of FRT for any of the following reasons:
(1)To assist in identifying a person that the officer has reasonable suspicion to believe has committed a felony.
(2)To assist in identifying any person who is deceased or who has been reported missing pursuant to Section 14211.
(3)To assist in identifying any person who has been lawfully arrested, during the process of booking or during that persons custodial detention.
(4)To assist in identifying any person if a peace officer determines that an emergency situation exists that involves immediate danger of death or serious physical injury to any person and the identification of that person is necessary to prevent that death or injury.
(b)If a peace officer uses or requests the use of FRT pursuant to subdivision (a), the officer shall document all of the following information:
(1)The identity of the peace officer using or requesting the use of FRT, and, if applicable, the officer authorizing the use or request.
(2)A detailed description, as available, of the person being identified.
(3)Any photograph or video being used as the probe image.
(4)Any details regarding other investigative measures taken to identify the person and an explanation of why those measures failed or are reasonably unlikely to succeed.
(c)(1)Beginning on July 1, 2024, and every six months thereafter, the custodian of any arrest photograph database being used in conjunction with FRT shall remove from the database any of the following images:
(A)Any photograph of a person under 18 years of age.
(B)Any photograph of an arrested or detained person who has been released without being charged with an offense, or whose charges have been dismissed.
(C)Any photograph of an arrested or detained person who has subsequently been acquitted of the charged offense.
(D)Any photograph of a person whose conviction has been expunged, has been exonerated of the crime, or who has had their conviction reversed on appeal.
(2)This section applies only to the use of a reference photograph database for the use of FRT and shall not be construed to prohibit a peace officer from using any other investigative database, including a fingerprint database.
(3)Any agency that maintains and operates an arrest photograph database shall establish procedures to ensure compliance with this subdivision.
13675.2.A peace officer using or requesting the use of FRT shall not do any of the following:
(a)Use FRT to identify any person solely on the basis that the person is exercising rights guaranteed by the United States Constitution, including free assembly, association, and speech.
(b)Rely on actual or perceived race, ethnicity, national origin, religion, disability, gender, gender identity, or sexual orientation in selecting a person to identify using FRT, except when there is trustworthy information, relevant to the locality and timeframe, in the context of a particular area and for a particular period of time, that links a person with a particular characteristic described in this subdivision to an identified criminal incident or scheme.
(c)Share FRT data with any state or federal agency for the purpose of enforcing federal immigration law.
(d)(1)Provide the results of, or information derived from, the use of FRT to any individual or to any agency or department in another state regarding the provision of lawful gender-affirming health care or gender-affirming mental health care performed in this state.
(2)As used in this subdivision, gender affirming health care and gender affirming mental health care have the same meaning as in Section 16010.2 of the Welfare and Institutions Code.
(e)(1)Provide the results of, or information derived from, the use of FRT to any individual or agency or department in another state regarding the provision of abortion services in this state.
(2)For purposes of this subdivision and subdivision (d), information derived from the use of FRT means information that would not have been discovered or obtained but for the use of FRT, regardless of any claim that the information would inevitably have been discovered or obtained through other means.
(f)Use an FRT match as the sole basis upon which probable cause is established for a search, arrest, or affidavit for a warrant. Any peace officers using information obtained from the use of FRT shall examine results with care and consider the possibility that matches could be inaccurate.
(g)Use FRT in conjunction with any reference photograph database that contains information, including images, obtained by any of the following means:
(1)In a manner that violates federal or state law.
(2)In a manner that violates a service agreement between a provider of an electronic communication service to the public or a provider of a remote computing service and customers or subscribers of that provider.
(3)In a manner that is inconsistent with the privacy policy of a provider described in paragraph (2), if applicable.
(4)By deceiving a person whose information was obtained.
(5)Through the unauthorized access of an electronic device or online account.
(6)In violation of a contract, court settlement, or other binding legal agreement.
(7)From unlawful or unconstitutional practices by any governmental official or entity.
13675.3.(a)By no later than January 31, 2025, and annually thereafter, each law enforcement agency using FRT, including agencies whose officers use or request the use of the FRT of another agency, shall prepare and submit a report to the California State Auditor containing only the following information regarding the use of FRT, as applicable:
(1)The information required to be documented by subdivision (b) of Section 13675.1 for each use or request for facial recognition.
(2)Whether modifications were made to any probe images and what those modifications were.
(3)The arrests that FRT results contributed to, and the offenses for which the arrests were made, disaggregated by race, ethnicity, gender, and age.
(4)A description of the reference photograph database that was used.
(5)A description of the FRT system that was used.
(6)The total number of searches performed.
(7)The total number of times FRT results were shared with another law enforcement agency, an outside agency, or a third party.
(8)Actions taken to comply with subdivision (c) of Section 13675.1 regarding unusable arrest photographs.
(b)By no later than January 31, 2025, and annually thereafter, each district attorneys office, city prosecutors office, and the Attorney General shall report to the California State Auditor the following information regarding information obtained from FRT:
(1)The number of convictions that FRT results contributed to and the offenses for which the convictions were obtained, disaggregated by race, ethnicity, gender, and age.
(2)The number of motions to suppress made related to FRT results, and the number granted or denied.
(c)On or before July 1 of each year, the California State Auditor shall release to the public, post online, and transmit to the Legislature a full and complete report concerning the use of FRT by law enforcement agencies and prosecutors, including any violations of this title that were identified.
(d)Any law enforcement agency using FRT shall keep and maintain those activity logs or other records necessary to comply with this section.
13675.4.(a)(1)A law enforcement agency shall not operate an FRT system that has not been evaluated under the National Institute of Standards and Technology Face Recognition Vendor Testing Program and achieved an accuracy score of 98 percent true positives within two or more datasets relevant to investigative applications on a program report.
(2)A peace officer shall not use FRT in any manner that reduces the facial recognition programs competency score below 98 percent.
(b)A law enforcement agency that uses FRT shall have a written policy that includes, without limitation, all of the following:
(1)A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT by the Commission on Peace Officer Standards and Training.
(2)A requirement that a manager authorized to use FRT be assigned to oversee the FRT program.
(3)A policy that describes the parameters of acceptable inputs to be used as probe images and that prohibits the use of sketches or other manually produced images.
(4)An acceptable use policy that includes specific allowances and restrictions on use.
(c)By no later than July 1, 2024, each law enforcement agency using FRT shall post the written policy required by subdivision (b) on its internet website.
(d)No law enforcement agency or peace officer shall request or enter into an agreement with another law enforcement agency or other third party to perform FRT searches on behalf of the requesting officer or agency if the program operated by the other agency or party does not meet the requirements of this section.
(e)(1)If the California State Auditor, pursuant to Section 13675.3, identifies any violations of this title by a law enforcement agency, that law enforcement agency shall cease using FRT until all violations have been corrected.
(2)If the use of FRT is suspended pursuant to this subdivision, the law enforcement agency shall notify the public of the suspension.
13675.5.(a)A law enforcement agency that uses FRT to attempt to identify an individual who is arrested shall provide to the individual both of the following:
(1)A notice of the name of the law enforcement agency that operated the FRT system used, and the name of the database, if any, that was used to identify the individual.
(2)A copy of the accuracy or bias report required by subdivision (a) of Section 13675.4, each probe image that was used by the agency, any modifications made to the probe image, the candidate list, in rank order, produced by the facial recognition system, and any other documentation related to the use of FRT in the law investigation.
(b)The information required under subdivision (a) shall be provided to the person in an appropriate language for the person if they are not fluent or literate in English.
(c)As used in this section candidate list means the top images that a facial recognition system determines to most closely match a probe image.
13675.6.(a)If a court or law enforcement agency determines that a peace officer has violated any provision of this title, and the court or agency finds that the circumstances surrounding the violation raise serious questions about whether or not the officer acted intentionally with respect to the violation, the agency shall promptly initiate a proceeding to determine whether disciplinary action against the officer is warranted.
(b)Notwithstanding any other law, a violation of this title is not punishable as a crime.
(c)A person who is subject to identification or attempted identification through FRT in violation of this title may bring a civil action against the peace officer or law enforcement agency responsible for the violation.
(d)A person who is the subject of disparate treatment or adverse impact on the basis of race, ethnicity, gender, or age, whether individually or as a member of a class of individuals, due to use of FRT or any technological element, criteria, method, or design feature thereof, by a law enforcement agency, may bring a cause of action against the peace officer, law enforcement agency, or maker of the facial recognition or face surveillance technology responsible for the violation.
(e)The following relief may be recovered or obtained in an action under subdivision (c) or (d):
(1)Any actual damages sustained by that person as a result of the violation.
(2)The greater of either of the following:
(A)Statutory damages of fifty thousand dollars ($50,000) per violation.
(B)Profits earned as a result of each violation.
(3)Exemplary damages pursuant to Section 3294 of the Civil Code.
(4)Injunctive, equitable, or declaratory relief as may be appropriate, including preliminary injunctive relief.
(5)Costs of the action, together with reasonable attorneys fees.
(f)A good faith reliance on a statutory authorization, or a good faith determination that the conduct complained of is permitted under this title, is a complete defense against any civil action brought under subdivision (c) or (d).
(g)A civil action under subdivision (b) or (c) may not be commenced later than two years after the date upon which the claimant discovered or first had a reasonable opportunity to discover the violation.
SEC. 2.If the Commission on State Mandates determines that this act contains costs mandated by the state, reimbursement to local agencies and school districts for those costs shall be made pursuant to Part 7 (commencing with Section 17500) of Division 4 of Title 2 of the Government Code.
SECTION 1.Section 13680 is added to the Penal Code, to read:13680.(a)Commencing July 1, 2024, a law enforcement agency that operates facial recognition technology (FRT) shall meet both of the following requirements:(1)Only FRT systems with algorithms that have been evaluated under the National Institute of Standards and Technology Face Recognition Vendor Test Program and have demonstrated an accuracy score of at least 98 percent true positives within two or more datasets relevant to investigative applications on a program report shall be used.(2)The agency shall have a written policy that includes all of the required provisions prescribed in subdivision (b).(b)The written policy required by subdivision (a) shall include, without limitation, all of the following:(1)A requirement that FRT use be limited to specifically authorized personnel who have received POST-certified training in the use of FRT.(2)A requirement that a manager be assigned to oversee the FRT program.(3)A policy that describes the parameters of acceptable inputs to be used for queries of available databases and that prohibits the use of sketches or other manually produced images.(4)An acceptable use policy that includes, without limitation, specific allowances and restrictions on use for all of the following:(A)To identify a suspect in alleged criminal behavior where reasonable suspicion exists that a crime has been or is being committed and the person whose image is being analyzed is the person who has committed or is committing that crime.(B)To identify a victim or witness to a crime.(C)To identify an unidentified deceased person.(D)To identify a person who is missing, incapacitated, or unable to identify themselves.(E)To identify a person who is lawfully detained and has not produced valid identification.(F)To investigate a credible threat of violence.(G)To mitigate an imminent threat to public safety.(5)A prohibited use policy that prohibits, without limitation, both of the following:(A)The use of FRT to identify an individual solely because of their race, color, religious beliefs, sexual orientation, gender, disability, national origin or membership in any other class protected by law against discrimination.(B)The use of FRT to identify an individual solely engaged in the exercise of a constitutionally protected right, including, without limitation, speech, public assembly, or the practice of religion, when the person has not individually violated any law, unless necessary to identify a victim or witness of a serious or violent felony, or to defend against an imminent or immediate threat to death or serious bodily injury.(6)A requirement that a record of all FRT queries be maintained by the agency.(c)FRT is an investigative tool to be used as an aid in identifying persons and generating investigative leads. An FRT-generated match shall not, under any circumstances, provide the sole basis for probable cause for an arrest, search, or affidavit for a warrant.(d)Any law enforcement agency that uses FRT shall post both of the following on their internet website:(1)A copy of the FRT policy described in subdivision (b).(2)Commencing on January 1, 2025, and annually thereafter, an annual report summarizing FRT usage for the previous calendar year.(e)(1)The admissibility or exclusion of an FRT query result as evidence in any court proceeding shall be governed by Chapter 4 (commencing with Section 305) of Division 2 of the Evidence Code.(2)This subdivision is declarative of existing law.(f)As used in this section, the following terms have the following meanings:(1)Facial recognition technology or FRT means a system that compares an input image of an unidentified human face against a database of known persons and, based on biometric data, generates possible matches to aid in identifying the person in the input image. For purposes of this section, FRT does not include any access control system used by a law enforcement agency that uses biometric inputs to confirm the identity of employees or other approved persons for the purpose of controlling access to any secured place, device, or system.(2)Law enforcement agency means any department or agency of the state or any political subdivision thereof that employs any peace officer as described in Section 830.(a)Commencing July 1, 2024, a law enforcement agency that operates facial recognition technology (FRT) shall meet both of the following requirements:
(1)Only FRT systems with algorithms that have been evaluated under the National Institute of Standards and Technology Face Recognition Vendor Test Program and have demonstrated an accuracy score of at least 98 percent true positives within two or more datasets relevant to investigative applications on a program report shall be used.
(2)The agency shall have a written policy that includes all of the required provisions prescribed in subdivision (b).
(b)The written policy required by subdivision (a) shall include, without limitation, all of the following:
(1)A requirement that FRT use be limited to specifically authorized personnel who have received POST-certified training in the use of FRT.
(2)A requirement that a manager be assigned to oversee the FRT program.
(3)A policy that describes the parameters of acceptable inputs to be used for queries of available databases and that prohibits the use of sketches or other manually produced images.
(4)An acceptable use policy that includes, without limitation, specific allowances and restrictions on use for all of the following:
(A)To identify a suspect in alleged criminal behavior where reasonable suspicion exists that a crime has been or is being committed and the person whose image is being analyzed is the person who has committed or is committing that crime.
(B)To identify a victim or witness to a crime.
(C)To identify an unidentified deceased person.
(D)To identify a person who is missing, incapacitated, or unable to identify themselves.
(E)To identify a person who is lawfully detained and has not produced valid identification.
(F)To investigate a credible threat of violence.
(G)To mitigate an imminent threat to public safety.
(5)A prohibited use policy that prohibits, without limitation, both of the following:
(A)The use of FRT to identify an individual solely because of their race, color, religious beliefs, sexual orientation, gender, disability, national origin or membership in any other class protected by law against discrimination.
(B)The use of FRT to identify an individual solely engaged in the exercise of a constitutionally protected right, including, without limitation, speech, public assembly, or the practice of religion, when the person has not individually violated any law, unless necessary to identify a victim or witness of a serious or violent felony, or to defend against an imminent or immediate threat to death or serious bodily injury.
(6)A requirement that a record of all FRT queries be maintained by the agency.
(c)FRT is an investigative tool to be used as an aid in identifying persons and generating investigative leads. An FRT-generated match shall not, under any circumstances, provide the sole basis for probable cause for an arrest, search, or affidavit for a warrant.
(d)Any law enforcement agency that uses FRT shall post both of the following on their internet website:
(1)A copy of the FRT policy described in subdivision (b).
(2)Commencing on January 1, 2025, and annually thereafter, an annual report summarizing FRT usage for the previous calendar year.
(e)(1)The admissibility or exclusion of an FRT query result as evidence in any court proceeding shall be governed by Chapter 4 (commencing with Section 305) of Division 2 of the Evidence Code.
(2)This subdivision is declarative of existing law.
(f)As used in this section, the following terms have the following meanings:
(1)Facial recognition technology or FRT means a system that compares an input image of an unidentified human face against a database of known persons and, based on biometric data, generates possible matches to aid in identifying the person in the input image. For purposes of this section, FRT does not include any access control system used by a law enforcement agency that uses biometric inputs to confirm the identity of employees or other approved persons for the purpose of controlling access to any secured place, device, or system.
(2)Law enforcement agency means any department or agency of the state or any political subdivision thereof that employs any peace officer as described in Section 830.
SECTION 1.It is the intent of the Legislature to enact legislation relating to the use of facial recognition technology by law enforcement agencies.It is the intent of the Legislature to enact legislation relating to the use of facial recognition technology by law enforcement agencies.
|
Date
|
Status
|
|---|---|
| Feb 1, 2024 | From committee: Filed with the Chief Clerk pursuant to Joint Rule 56. |
| Jan 31, 2024 | Died pursuant to Art. IV, Sec. 10(c) of the Constitution. |
| May 18, 2023 | In committee: Held under submission. |
| May 18, 2023 | Joint Rule 62(a), file notice suspended. (Page 1580.) |
| May 17, 2023 | In committee: Set, first hearing. Referred to APPR. suspense file. |
| May 2, 2023 | Re-referred to Com. on APPR. |
| May 1, 2023 | Read second time and amended. |
| Apr 27, 2023 | From committee: Amend, and do pass as amended and re-refer to Com. on APPR. (Ayes 9. Noes 0.) (April 25). |
| Apr 20, 2023 | Re-referred to Com. on P. & C.P. |
| Apr 19, 2023 | From committee chair, with author's amendments: Amend, and re-refer to Com. on P. & C.P. Read second time and amended. |
| Apr 12, 2023 | From committee: Do pass and re-refer to Com. on P. & C.P. (Ayes 6. Noes 0.) (April 11). Re-referred to Com. on P. & C.P. |
| Apr 5, 2023 | Re-referred to Com. on PUB. S. |
| Apr 4, 2023 | From committee chair, with author's amendments: Amend, and re-refer to Com. on PUB. S. Read second time and amended. |
| Mar 28, 2023 | In committee: Set, first hearing. Hearing canceled at the request of author. |
| Mar 6, 2023 | Re-referred to Com. on PUB. S. |
| Mar 2, 2023 | From committee chair, with author's amendments: Amend, and re-refer to Com. on PUB. S. Read second time and amended. |
| Mar 2, 2023 | Referred to Coms. on PUB. S. and P. & C.P. |
| Feb 10, 2023 | From printer. May be heard in committee March 12. |
| Feb 9, 2023 | Read first time. To print. |
|
Date
|
Description
|
Location
|
Votes
|
|
|---|---|---|---|---|
| Apr 25, 2023 | Do pass as amended and be re-referred to the Committee on [Appropriations] | Assembly Standing Committee on Privacy and Consumer Protection |
Passed |
|
| Apr 11, 2023 | Do pass and be re-referred to the Committee on [Privacy and Consumer Protection] | Assembly Standing Committee on Public Safety |
Passed |
This list shows positions taken on this bill by organizations on the date displayed, which is linked to the original source.
Positions by individuals are not included.
|
Organization
|
Position
|
Date
|
|---|---|---|
| Access Reproductive Justice | ** Against | |
| Aclu California Action | ** Against | |
| All of Us or None - Riverside | ** Against | |
| A New Way of life | ** Against | |
| Anti Police-Terror Project | ** Against | |
| Arcadia Police Officers' Association | ** Support | |
| Asian Americans Advancing Justice - Asian Law Caucus | ** Against | |
| Asian Law Alliance | ** Against | |
| Bend the Arc California | ** Against | |
| Bend The Arc: Jewish Action | ** Against | |
| Burbank Police Officers' Association | ** Support | |
| California Association of Black Lawyers | ** Against | |
| California Coalition for Women Prisoners | ** Against | |
| California Coalition of School Safety Professionals | ** Support | |
| California Immigrant Policy Center | ** Against | |
| California Innocence Coalition: Northern California Innocence Project, California Innocence Project, Loyola Project for The Innocent | ** Against | |
| California Public Defenders Association | ** Against | |
| California United for a Responsible Budget | ** Against | |
| Cancel the Contract | ** Against | |
| Care First California | ** Against | |
| Central American Resource Center of San Francisco | ** Against | |
| Citizens for A Better Los Angeles | ** Against | |
| CLAREMONT POLICE OFFICERS ASSOCIATION | ** Support | |
| Clergy and Laity United for Economic Justice | ** Against | |
| Coalition for Homelessness San Francisco | ** Against | |
| Corona Police Officers Association | ** Support | |
| Council on American Islamic Relations | ** Against | |
| Culver City Police Officers' Association | ** Support | |
| Deputy Sheriffs Association of Monterey County | ** Support | |
| Electronic Frontier Foundation | ** Against | |
| Equal Justice Society | ** Against | |
| Family Reunification Equity & Empowerment | ** Against | |
| Fight For The Future | ** Against | |
| Free Speech Coalition | ** Against | |
| Fullerton Police Officers' Association | ** Support | |
| If/when/how: Lawyering for Reproductive Justice | ** Against | |
| Indivisible Ca Statestrong | ** Against | |
| Initiate Justice | ** Against | |
| Lawyers Committee for Civil Rights of the San Francisco Bay Area | ** Against | |
| Lawyers' Committee for Civil Rights of the San Francisco Bay Area | ** Against | |
| League Of California Cities | ** Support | |
| Legal Services for Prisoners with Children | ** Against | |
| Long Beach Immigrant Rights Coalition | ** Against | |
| los angeles school police officers association | ** Support | |
| Media Alliance | ** Against | |
| Mediajustice | ** Against | |
| Murrieta Police Officers' Association | ** Support | |
| Muslim Democrats and Friends | ** Against | |
| National Center for Lesbian Rights | ** Against | |
| Newport Beach Police Association | ** Support | |
| Oakland Privacy | ** Against | |
| Orange County Board of Supervisors - Supervisor Vicente Sarmiento | ** Support | |
| Orange County Rapid Response Network | ** Against | |
| Palos Verdes Police Officers Association | ** Support | |
| Partnership for the Advancement of New Americans | ** Against | |
| People's Budget Orange County | ** Against | |
| Placer County Deputy Sheriffs' Assocation | ** Support | |
| Pomona Police Officers' Association | ** Support | |
| Positive Women's Network - USA | ** Against | |
| Privacy Rights Clearinghouse | ** Against | |
| RIVERSIDE POLICE OFFICERS ASSOCIATION | ** Support | |
| Riverside Sheriffs' Association | ** Support | |
| Safer Streets LA | ** Against | |
| San Francisco Public Defender - Racial Justice Committee | ** Against | |
| San Jose Nikkei Resisters | ** Against | |
| Santa Ana Police Officers Association | ** Support | |
| Secure Justice | ** Against | |
| Security Industry Association | ** Support | |
| Silicon Valley De-Bug | ** Against | |
| Starting Over Inc. | ** Against | |
| St. James Infirmary | ** Against | |
| Stop the Musick Coalition | ** Against | |
| Tenth Amendment Center | ** Against | |
| Training In Early Abortion For Comprehensive Healthcare | ** Against | |
| Transforming Justice Orange County | ** Against | |
| Transgender, Gendervariant, Intersex Justice Project | ** Against | |
| Upland Police Officers Association | ** Support | |
| URGE: Unite for Reproductive & Gender Equity | ** Against |
Date of Hearing: May 17, 2023
ASSEMBLY COMMITTEE ON APPROPRIATIONS
Chris Holden, Chair
ABPCA Bill Id:AB 642
Author:(Ting) – As Amended Ver:May 1, 2023
Policy Committee: | Public Safety | Vote: | 6 - 0 |
Privacy and Consumer Protection | 9 - 0 | ||
|
|
Urgency: No State Mandated Local Program: Yes Reimbursable: Yes
SUMMARY:
This bill sets minimum standards for use of facial recognition technology (FRT) by law enforcement agencies.
Specifically, this bill:
- Authorizes a peace officer to use or request the use of FRT for specified reasons, including to assist in identifying a person that the officer has reasonable suspicion to believe has committed a felony.
- Requires an officer who uses or requests the use of FRT to document the image depicting the person to be identified, a detailed description of that person, and any details about investigative measures taken to identify the person and an explanation of why those measures failed or are reasonably unlikely to succeed.
- Requires the custodian of an arrest photograph database used in conjunction with FRT to remove certain images every six months, beginning in July 2024, including images of minors, people who have been acquitted, and people who were released without being charged or had their charges dismissed.
- Prohibits a law enforcement agency or officer from using FRT in certain circumstances, including using FRT to identify someone on the sole basis that they are exercising their constitutional right to assembly or free speech, releasing FRT data or results in a way that violates the California Values Act or someone’s exercise of their right to access abortion, or using FRT with a reference photograph database that contains images obtained illegally or in violation of contract.
- Requires a law enforcement agency that uses FRT to report annually to the State Auditor, beginning in January 2025, specified details about each instance of the agency’s use of FRT.
- Requires public prosecutors to report annually to the State Auditor, beginning in January 2025, about convictions obtained using FRT and motions to suppress evidence related to FRT results.
- Requires the State Auditor to annually release to the public and the Legislature a full and complete report about the use of FRT by law enforcement agencies and prosecutors, including any violations that were identified.
- Specifies that a law enforcement agency must stop using FRT until any violations identified by the State Auditor are remedied.
- Specifies that a law enforcement agency may only operate an FRT system that has been evaluated under the National Institute of Standards and Technology Face Recognition Vendor Testing Program and achieved a specified accuracy score.
- Requires the California Department of Technology (CDT) to issue standards by July 2024 to ensure the confidentiality and cybersecurity of FRT data and results.
- Requires a law enforcement agency that uses FRT to have a written policy that follows the CDT guidelines described in 10) and includes other mandatory provision, including that FRT may only be used by someone who has received certified training on FRT by the Commission on Peace Officer Standards and Training (POST).
- Requires a law enforcement agency to initiate a proceeding to determine whether disciplinary action is warranted against a peace officer if a court or the agency determines that the officer violated any of these provisions and the court or agency finds that the circumstances surrounding the violation raise serious questions about whether or not the officer acted intentionally with respect to the violation.
- Specifies that a violation of these provisions is not punishable as a crime.
- Allows a person to bring a civil action against a peace officer or law enforcement agency responsible if the person is subject to identification or attempted identification through FRT in violation of these provisions.
- Allows a person to bring a civil action against a peace officer, law enforcement agency, or maker of facial recognition or face surveillance technology if the person is subject to disparate treatment or adverse impact on the basis of race, ethnicity, gender, or age resulting from use of FRT.
- Allows a person bringing a lawsuit under 14) or 15) to recover actual damages, exemplary damages, injunctive equitable, or declaratory relief, costs and attorney’s fees, and the greater of either $50,000 in statutory damages or the profits earned as a result of each violation.
FISCAL EFFECT:
- One-time costs (General Fund (GF)) to POST in the low hundreds of thousands of dollars to develop training on FRT. This bill requires that any officer who uses FRT must first complete an FRT training by POST. POST reports that it does not have existing training on this subject. POST may also experience ongoing costs in administering the training to officers and updating it as the technology changes and develops.
- Absorbable costs (GF) to CDT to issue standards to ensure the confidentiality and cybersecurity of FRT data and results. CDT anticipates only minor costs for generating the standards required by this bill.
- Costs (GF) to the State Auditor of an unknown, but potentially significant, amount to produce an annual report about the use of FRT by law enforcement agencies and prosecutors. Due to the volume of information reported to the Auditor under this bill, and the requirement that the Auditor identify any violations within that information, costs to the Auditor may exceed $150,000 annually.
- Costs (GF) to the Attorney General and potentially reimbursable costs (local funds, GF) to district attorneys’ and city prosecutors’ offices to report to the State Auditor information about convictions and motions to suppress relating to information obtained by FRT. It is unclear whether public prosecutors already collect this information. To the extent local prosecutors incur costs in fulfilling the bill’s reporting requirements, General Fund costs will depend on whether these duties constitute a reimbursable state mandate, as determined by the Commission on State Mandates.
- Possibly non-reimbursable costs (local funds, GF) to local law enforcement agencies that use FRT to complete FRT trainings, draft FRT policies, create and maintain records of FRT use, and report annually about FRT use to the State Auditor. GF costs will depend on whether the duties imposed by this bill constitute a reimbursable state mandate, as determined by the Commission on State Mandates.
- Costs to the courts (Trial Court Trust Fund, GF) of an unknown but potentially significant amount to adjudicate civil litigation resulting from this bill. The bill creates two new private rights of action that allow specified individuals to bring civil litigation if they experience specified harms resulting from FRT. An average court day costs $8,000, so if 10 individuals file lawsuits due to this bill and each suit requires two days of court time, the resulting cost would be roughly $160,000. Although courts are not funded on the basis of workload, increased pressure on the Trial Court Trust Fund may create a need for increased funding for courts from the General Fund.
COMMENTS:
- Author’s Statement. According to the author:
I authored AB 1215 in 2019 which banned the use of biometric surveillance through police body cameras. The bill only passed with a three year moratorium that expired January 1, 2023. Consequently, current law has absolutely no parameters or restrictions regarding law enforcement’s use of facial recognition technology. It is critical that we ensure there are safeguards in place in order to avoid another year of unregulated use. We can’t go another year with no protections. AB 642 is a response to a battle that we cannot afford to risk losing. The bill includes critical safeguards such as codifying an accuracy level, accountability and oversight, reporting, civil penalties, protections for immigrant communities, people seeking services such as abortion and gender affirming care, prohibits law enforcement from using a match alone to arrest someone, to request a warrant, to violate someone’s constitutional rights, and to discriminate against protected characteristics. Most importantly, this bill does not prohibit nor deter local governments from choosing to ban the use of facial recognition technology.
- Facial Recognition Technology. FRT, as defined in this bill, is a system that compares a probe image of an unidentified human face against a reference photograph database, and, based on biometric data, generates possible matches to aid in identifying the person in the probe image. More colloquially, FRT is a way of identifying a person or confirming their identity using an image of their face. In the context contemplated by this bill, a peace officer can compare a photograph of a person the officer wants to identify against an FRT database that contains images of many people. If the FRT system identifies a matching image in the database, the officer can use the personal information associated with the database image to identify the person in the officer’s photograph.
There are legitimate concerns about the potential misuse of FRT and the proliferation of biometric surveillance. FRT is a powerful and potentially intrusive tool, and, as with similar contemporary technologies, its use by law enforcement may have a disproportionate negative impact on people of color and other historically disadvantaged groups. As described in detail in the analyses of this bill by the Assembly Committee on Public Safety and the Assembly Committee on Privacy and Consumer Protection, use of this emerging technology by law enforcement has serious implications for the legal rights, civil liberties, and privacy of individual Californians.
The bill establishes minimum use standards for law enforcement agencies that deploy FRT, requires CDT to create data use guidelines to ensure the security of FRT data and requires law enforcement agencies to adhere to CDT’s guidelines, and creates private right of actions for individual citizens to attempt to enforce certain violations. The bill also requires law enforcement agencies and prosecutors to report information about their use of FRT to the State Auditor, and requires the State Auditor to report publicly and to the Legislature about violations committed by those groups.
- Related Legislation. AB 1034 (Wilson) prohibits a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other officer camera. AB 1034 is pending on the Assembly floor.
AB 793 (Bonta) prohibits a government entity from seeking or enforcing a reverse-location demand or reverse-keyword demand, as defined. AB 793 is pending on this committee’s suspense file.
Prior Legislation. AB 1215 (Ting), Chapter 579, Statutes of 2019, prohibited a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other camera. These prohibitions expired on January 1, 2023 at the law’s sunset date.
SB 1038 (Bradford), of the 2021-22 Legislative Session, would have deleted the sunset date on provisions enacted by AB 1215. SB 1038 died on the inactive file in the Senate.
Analysis Prepared by: Annika Carlson / APPR. / (916) 319-2081
Date of Hearing: April 25, 2023
ASSEMBLY COMMITTEE ON PRIVACY AND CONSUMER PROTECTION
Jesse Gabriel, Chair
ABPCA Bill Id:AB 642 (
Author:Ting) – As Amended Ver:April 19, 2023
SUBJECT: Law enforcement agencies: facial recognition technology
SYNOPSIS
This bill proposes to regulate the use of facial recognition technology (FRT) by local law enforcement agencies. According to the author, the bill includes critical safeguards such as codifying an accuracy level; accountability and oversight; reporting; civil penalties; protections for immigrant communities and people seeking services such as abortion and gender affirming care; prohibiting law enforcement from using a match alone to arrest someone or to request a warrant; and protections against violating someone’s constitutional right and discriminating on the basis of protected characteristics. Unfortunately, this bill does not yet include all of the safeguards the author is hoping for.
FRT technology remains far from perfect. Recent studies continue to highlight that many FRT systems continue to be less effective at identifying people of color, women, older people, and children, even when the technology is being tested in optimal conditions using clear probe photographs. Tests are not performed in the real world, where police routinely conduct searches using real world images—which are frequently blurry and/or distant—producing bad results that are even more likely to be mismatched. The controversy surrounding law enforcement use of facial recognition has led many California cities to ban the technology, including San Francisco, Oakland, Berkeley, Santa Cruz, and Alameda. Despite the ban in San Francisco, officers there may have skirted the city’s ban by outsourcing an FRT search to another law enforcement agency.
Developing a workable regulatory framework that acknowledges the utility of FRT for safety, security, and efficiency, while remaining conscientious of the potential for FRT to infringe on fundamental rights and civil liberties, such as the individual right to privacy and the freedom to express viewpoints anonymously, is indeed difficult, and requires consideration of several critical questions, both practical and conceptual. Further exacerbating the complexity of this task is the necessity in such a framework for sensitivity to the current technical shortcomings of FRT, including performance disparities between demographic groups and entrenchment of existing cultural biases by those designing and training the algorithms underlying these technologies.
The question before this Committee is whether or not this bill furthers the Committee’s policy priorities. First and foremost, protecting Californians’ constitutional right to privacy. The Committee is also working to ensure that all Californians, and those coming from out of state, are protected from punitive and discriminatory draconian laws attacking the LGBTQ+ community and criminalizing people seeking abortion and gender affirming care. Another significant priority of the Committee is ensuring that the State’s laws protect our immigrant neighbors from federal policies that make them vulnerable to being separated from their families, imprisoned, and ultimately returned to countries that many were forced to flee for their own safety.
As the following excerpt from Robert Williams’ letter of opposition articulates, there is current trauma being inflicted on men like Mr. Williams and their families by use of FRT. If the Legislature decides that the best path is regulation, as the author asserts, it needs to be done through a slow, deliberative, thoughtful process that includes a broad range of stakeholders and experts.
“I may be the first documented case of a wrongful arrest based on a false face recognition match, but I wasn’t the last. In the years following my arrest, police have similarly misidentified and detained four other individuals we know of across the country. All of us are Black men. All of us suffered the trauma of being wrenched from our families, isolated, and interrogated by police officers who dismissed our claims of innocence because they believed the technology was infallible. In fact, studies have found that Black and Asian people are up to 100 times more likely to be misidentified than white men.” – Robert Williams
This bill is author sponsored and has two supporters and over 30 civil rights and social justice organizations in opposition.
This bill previously passed the Assembly Public Safety Committee on a 6-0-2 vote.
SUMMARY: Sets minimum standards for use of facial recognition technology (FRT) by law enforcement, including requiring law enforcement agencies to have a written policy for FRT use, allowing for FRT use when a peace officer has reasonable suspicion that an individual has committed a felony, and providing that an FRT-generated match of an individual may not be the sole basis for probable cause for an arrest, search, or affidavit for a warrant. Specifically, this bill:
- Defines “facial recognition technology” or “FRT” to mean a system that compares a probe image of an unidentified human face against a reference photograph database, and, based on biometric data, generates possible matches to aid in identifying the person in the probe image.
- FRT includes any surveillance system that actively uses FRT to identify persons in a surveilled area in real time.
- FRT does not include any access control system used by a law enforcement agency that uses biometric inputs to confirm the identity of employees or other approved persons for the purpose of controlling access to any secured place, device, or system.
- Defines “reference photograph database” to mean a database populated with photographs of individuals who have been identified, including databases composed of driver’s licenses or other documents made or issued by or under the authority of the state, a political subdivision thereof, databases operated by third parties, and arrest photograph databases.
- Defines “arrest photograph database” to mean a database populated primarily by booking or arrest photographs or other photographs of persons with law enforcement contacts.
- Defines “probe image” to mean an image of a person that is searched against a database of known, identified persons or an unsolved photograph file.
- Authorizes a peace officer to use or request the use of FRT for any of the following reasons:
- To assist in identifying a person that officer has reasonable suspicion to believe has committed a felony.
- To assist in identifying a person who is deceased or who has been reported missing, as specified.
- To assist in identifying any person who has been lawfully arrested, during the process of booking or during that person’s custodial detention.
- To assist in identifying any person if a peace officer determines that an emergency situation exists that involves immediate danger of death or serious physical injury to any person and the identification of that person is necessary to prevent that death or injury.
- Requires, if a peace officer uses or requests the use of FRT, documentation of all of the following information:
- The identity of the peace officer using or requesting the use of FRT, and, if applicable, the officer authorizing the use or request;
- A detailed description, as available, of the person being identified;
- Any photograph or video being used as a probe image; and,
- Any details regarding other investigative measures taken to identify the person and an explanation of why those measures failed or are reasonably unlikely to succeed.
- Requires the custodian of any arrest photograph database being used in conjunction with FRT, beginning on July 1, 2024 and every six months thereafter, to remove from the database any of the following images:
- Any photograph of a person under 18 years of age;
- Any photograph of an arrested or detained person who has been released without being charged with an offense, or whose charges have been dismissed;
- Any photograph of an arrested or detained person who has subsequently been acquitted of the charged offense; or,
- Any photograph of a person whose conviction has been expunged, has been exonerated of the crime, or who has had their conviction reversed on appeal.
- Provides that the image removal requirement applies only to the use of a reference photograph database for the use of FRT and shall not be construed to prohibit a peace officer from using any other investigative database including a fingerprint database.
- Requires any agency that maintains and operates an arrest photograph database to establish procedures to ensure compliance with the image removal requirement.
- Prohibits a peace officer using or requesting the use of FRT from doing any of the following:
- Using FRT to identify any person solely on the basis that the person is exercising rights guaranteed by the United States Constitution, including free assembly, association, and speech;
- Relying on actual or perceived race, ethnicity, national origin, religion, disability, gender, gender identity, or sexual orientation in selecting a person to identify using FRT, except when there is trustworthy information, relevant to the locality and timeframe, in the context of a particular area and for a particular period of time, that links a person with a particular characteristic described to an identified criminal incident or scheme;
- Sharing FRT data with any state or federal agency for the purpose of enforcing federal immigration law;
- Providing the results of, or information derived from, the use of FRT to any individual or to any agency or department in another state regarding the provision of lawful gender-affirming health care or gender-affirming mental health care performed in this state;
- Providing the results of, or information derived from the use of FRT to any individual or agency or department in another state regarding the provision of abortion services in this state;
- Using an FRT match as the sole basis upon which probable cause is established for a search, arrest, or affidavit for a warrant, and provides that any peace officer using information obtained from the use of FRT shall examine results with care and consider the possibility that matches could be inaccurate; and
- Using FRT in conjunction with any reference photograph database that contains information, including images, obtained by any of the following means:
- In a manner that violates federal or state law;
- In a manner that violates a service agreement between a provider of an electronic communication service to the public or a provider of a remote computing service and customers or subscribers of that provider;
- In a manner that is inconsistent with the privacy policy of a provider, as specified;
- By deceiving a person whose information was obtained;
- Through the unauthorized access of an electronic device or online account;
- In violation of a contract, court settlement, or other binding legal agreement; or
- From unlawful or unconstitutional practices by any governmental official or entity.
- Provides that, for the purpose of prohibiting sharing FRT results regarding persons receiving abortion services or gender-affirming health care and gender-affirming mental health care, “information derived from the use of FRT” means information that would not have been discovered or obtained but for the use of FRT, regardless of any claim that the information would inevitably have been discovered or obtained through other means.
- Requires each law enforcement agency using FRT or requesting the use of FRT of another agency, by no later than January 31, 2025, and annually thereafter, to prepare and submit a report to the California State Auditor containing only the following information regarding the use of FRT, as applicable:
- The information a peace officer is required to document when using or requesting the use of FRT;
- Whether modifications were made to any probe images and what those modifications were;
- The arrests that FRT results contributed to, and the offenses for which the arrests were made, disaggregated by race, ethnicity, gender, and age;
- A description of the reference photograph database that was used;
- A description of the FRT system that was used;
- The total number of searches performed;
- The total number of times FRT results were shared with another law enforcement agency, an outside agency, or a third party; and
- The actions taken to comply with requirement that a law enforcement agency remove certain images from a reference photograph database, as specified.
- Requires each district attorney’s office, city prosecutor’s office, and the Attorney General, by no later than January 31, 2025 and annually thereafter, to report to the California State Auditor the following information regarding information obtained from FRT:
- The number of convictions that FRT results contributed to and the offenses for which the convictions were obtained, disaggregated by race, ethnicity, gender, and age; and
- The number of motion to suppress made related to FRT results, and the number granted or denied.
- Requires the State Auditor, on or before July 1 of each year, to release to the public, post online, and transmit to the Legislature, a full and complete report concerning the use of FRT by law enforcement agencies and prosecutors, including any violations identified.
- Requires any law enforcement agency using FRT to keep and maintain FRT activity logs or other required records, as specified.
- Prohibits a law enforcement agency from operating an FRT system that has not been evaluated under the National Institute of Standards and Technology Face Recognition Vendor Testing Program and achieved an accuracy score of 98 percent true positives within two or more datasets relevant to investigative applications on a program report.
- Prohibits a peace officer from using FRT in any manner that reduces the program’s competency score below 98 percent.
- Requires a law enforcement agency that uses FRT to have a written policy that includes, without limitation, all of the following:
- A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT by the Commission of Peace Officer Standards and Training;
- A requirement that a manger authorized to use FRT be assigned to oversee the FRT program;
- A policy that describes the parameters of acceptable inputs to be used as probe images and that prohibits the use of sketches or other manually produced images; and
- An acceptable use policy that includes specific allowances and restrictions on use.
- Requires each law enforcement agency using FRT, by no later than July 1, 2024, to post the required written policy on its website.
- Prohibits a law enforcement agency or peace officer from requesting or entering into an agreement with another law enforcement agency or other third party to perform FRT search on behalf of the requesting officer or agency if the program operated by the other agency or party does not meet the specified accuracy requirements.
- Requires a law enforcement agency, if the State Auditor identifies any violations by that agency, to cease using FRT until all violations have been corrected.
- Requires the law enforcement agency to notify the public if its use of FRT is suspended for violations, as specified.
- Requires a law enforcement agency that uses FRT to attempt to identify an individual who is arrested to provide the individual with both of the following:
- A notice of the name of the law enforcement agency that operated the FRT system used, and the name of the database, if any, the was used to identify the individual; and
- A copy of the required accuracy or bias report, each probe image that was used by the agency, any modifications made to the probe image, the candidate list, in rank order, produced by the facial recognition system, and any other documentation related to the use of FRT in the law investigation.
- Requires the notice provided by the law enforcement agency to the individual arrested to be an appropriate language for the person if they are not fluent or literate in English.
- Provides that, if a court or law enforcement agency determines that a peace officer has used FRT in violation of the law, and the court or agency finds that the circumstances surrounding the violation raise serious questions about whether or not the officer acted intentionally with respect to the violation, the agency shall promptly initiate a proceeding to determine whether disciplinary action against the officer is warranted.
- Provides that, notwithstanding any other law, a violation of this title is not punishable as a crime.
- Authorizes a person who is subject to identification or attempted identification through FRT in violation of the law to bring a civil action against the peace officer or law enforcement agency responsible for the violation.
- Provides that a person who is the subject of disparate treatment or adverse impact on the basis of race, ethnicity, gender, or age, whether individually or as a member of a class of individuals, due to use of FRT or any technological element, criteria, method, or design feature thereof, by a law enforcement agency, may bring a cause of action against the peace officer, law enforcement agency, or maker of the facial recognition or face surveillance technology responsible for the violation.
- Provides that the following relief may be recovered or obtained in a civil action for an FRT use violation:
- Any actual damages sustained by that person as a result of the violation;
- The greater of either of the following:
- Statutory damages of $50,000 per violation; or
- Profits earned as a result of each violation.
- Exemplary damages, as specified;
- Injunctive, equitable, or declaratory relief as may be appropriate, including preliminary injunctive relief; and,
- Costs of the action, together with reasonable attorney’s fees.
- Prohibits a civil action from commencing later than two years after the date upon which the claimant discovered or first had a reasonable opportunity to discover the violation.
EXISTING LAW:
- Provides, pursuant to the California Constitution, that all people have inalienable rights, including the right to pursue and obtain privacy. (Cal. Const., art. I, § 1.)
- Provides, pursuant to the Unruh Civil Rights Act, that all persons within the jurisdiction of this state are free and equal, and no matter what their sex, race, color, religion, ancestry, national origin, disability, medical condition, genetic information, marital status, sexual orientation, citizenship, primary language, or immigration status are entitled to the full and equal accommodations, advantages, facilities, privileges, or services in all business establishments of every kind whatsoever. (Civ. Code § 51.)
- Provides that no person in the State of California shall, on the basis of sex, race, color, religion, ancestry, national origin, ethnic group identification, age, mental disability, physical disability, medical condition, genetic information, marital status, or sexual orientation, be unlawfully denied full and equal access to the benefits of, or be unlawfully subjected to discrimination under, any program or activity that is conducted, operated, or administered by the state or by any state agency, is funded directly by the state, or receives any financial assistance from the state. (Gov. Code §§ 11135 et. seq.)
- Establishes the California Values Act, which prohibits state law enforcement from using state resources to assist in the enforcement of federal immigration law, except as specified. (Gov. Code § 7282 et seq.)
- Establishes California as a sanctuary state and prohibits any law enforcement agency from cooperating with federal immigration enforcement authorities. (Gov. Code § 7284, et seq.)
- Prohibits use of California state funds for travel to any state that is subject to a ban on state-funded and state-sponsored travel because that state enacted a law that voids or repeals, or has the effect of voiding or repealing, existing state or local protections against discrimination on the basis of sexual orientation, gender identity, or gender expression, or has enacted a law that authorizes or requires discrimination against same-sex couples or their families on the basis of sexual orientation, gender identity, or gender expression. (Gov. Code § 11139.8.)
- Establishes the Reproductive Privacy Act, which provides that the Legislature finds and declares that every individual possesses a fundamental right of privacy with respect to personal reproductive decisions, which entails the right to make and effectuate decisions about all matters relating to pregnancy, including prenatal care, childbirth, postpartum care, contraception, sterilization, abortion care, miscarriage management, and infertility care. Accordingly, it is the public policy of the State of California that:
- Every individual has the fundamental right to choose or refuse birth control.
- Every individual has the fundamental right to choose to bear a child or to choose to obtain an abortion, with specified limited exceptions.
- The state shall not deny or interfere with a person’s fundamental right to choose to bear a child or to choose to obtain an abortion, except as specifically permitted. (Health & Saf. Code § 123462.)
- Provides that the state may not deny or interfere with a person’s right to choose or obtain an abortion prior to viability of the fetus or when the abortion is necessary to protect the life or health of the person. (Health & Saf. Code § 123466 (a).)
- States that a person shall not be compelled in a state, county, city, or other local criminal, administrative, legislative, or other proceeding to identify or provide information that would identify or that is related to an individual who has sought or obtained an abortion if the information is being requested based on either another state’s laws that interfere with a person’s rights under subdivision (a) or a foreign penal civil action. (Health & Saf. Code § 123466(b).)
- Declares that it is the intent of the Legislature to establish policies and procedures to address issues related to the downloading and storage data recorded by a body-worn camera worn by a peace officer; these policies and procedures shall be based on best practices. (Pen. Code § 832.18(a).)
- Encourages agencies to consider best practices in developing policies related to the use of body-worn cameras and the storage of the data obtained from these cameras. (Pen. Code, § 832.18.)
- Instructs law enforcement agencies to work with legal counsel to determine a retention schedule to ensure that storage policies and practices are in compliance with all relevant laws and adequately preserve evidentiary chains of custody. (Pen. Code, § 832.18(b)(5)(D).)
- Instructs a law enforcement agency using a third-party vendor to manage its data storage system to consider the following factors to protect the security and integrity of the data: Using an experienced and reputable third-party vendor; entering into contracts that govern the vendor relationship and protect the agency’s data; using a system that has a built-in audit trail to prevent data tampering and unauthorized access; using a system that has a reliable method for automatically backing up data for storage; consulting with internal legal counsel to ensure the method of data storage meets legal requirements for chain-of-custody concerns; and using a system that includes technical assistance capabilities. (Pen. Code, § 832.18(b)(7).)
- Requires that a public agency that operates or intends to operate an Automatic License Plate Recognition (ALPR) system to provide an opportunity for public comment at a public meeting of the agency's governing body before implementing the program. (Civ. Code § 1798.90.55.)
FISCAL EFFECT: As currently in print this bill is keyed fiscal
COMMENTS:
1) Facial recognition technology (FRT) refers to the use of automated devices to identify or verify a person from a digital image by determining whether two images of faces represent the same person. FRT consists of two component processes: face detection, or locating a face within a photo, and face identification, or the matching of facial information to an image or images in a specified database that links to identifying information. FRT relies on the use of biometrics, the statistical analysis of measurements of biological data, in order to compare these images, reducing complex images to numerical values that represent key facial measurements that distinguish individuals.
2) Research demonstrates significant problems with FRT and its ability to accurately identify people. FRT technology remains far from perfect. Recent studies continue to highlight that many FRT systems are less effective at identifying people of color, women, older people, and children. These race, gender, and age biases arise because FRT is often “trained” using non-diverse faces. As a result, police relying on the technology to identify people have wrongfully arrested Black men based on mistaken FRT identifications, known as “false positives.”
Numerous studies reveal these FRT performance inconsistencies in identifying non-white males and people with darker complexions, generally. The National Institute of Standards and Technology (NIST) conducted the most prominent of these global studies. Their 2019 analysis of 189 facial recognition software programs found that people of color were up to 100 times more likely to be wrongfully identified than white men. (Johnson, et al, Facial recognition systems in policing and racial disparities in arrests, Government Information Quarterly 39 (2022) 101753, Elsevier, available at https://www.sciencedirect.com/science/article/abs/pii/S0740624X22000892.)
Clare Garvie, an expert in law enforcement use of FRT, notes that these NIST tests are performed in a controlled environment using clear images. They are not performed in the real world, where police routinely conduct searches using real world images–which are frequently blurry and/or distant–producing bad results that are even more likely to be mismatched by FRT. (Garvie, Garbage In, Garbage Out: Face Recognition on Flawed Data. Georgetown Law Center on Privacy and Technology, (May 16, 2019) available at https://www.law.georgetown.edu/privacy-technology-center/publications/garbage-in-garbage-out-face-recognition-on-flawed-data/.)
Not only does FRT have a racial bias problem, research shows that it also has a gender problem. One study, conducted by Colorado University at Boulder, found that with a brief glance, facial recognition software can categorize gender with remarkable accuracy. But if that face belongs to a transgender person, such systems get it wrong more than one-third of the time. In addition, earlier studies suggest software tends to be most accurate when assessing the gender of white men but misidentify women of color as much as one-third of the time.
According to the study’s lead author, Morgan Klaus Scheuerman, “We found that facial analysis services performed consistently worse on transgender individuals, and were universally unable to classify non-binary genders. While there are many different types of people out there, these systems have an extremely limited view of what gender looks like.”
The Colorado study suggests that FRT systems identify gender based on outdated stereotypes. When Scheuerman, a male with long hair, submitted his picture, half categorized him as female. ‘These systems run the risk of reinforcing stereotypes of what you should look like if you want to be recognized as a man or a woman,” said Scheuerman. “That impacts everyone.” (Facial recognition software has a gender problem, National Science Foundation (Nov. 1, 2019), available at https://new.nsf.gov/news/facial-recognition-software-has-gender-problem.)
3) Law Enforcement Uses of Facial Recognition Systems. Despite growing concerns, law enforcement agencies at the federal, state, and local level continue to use facial recognition programs. A recent Government Accountability Office report revealed that 20 federal agencies employ such programs, 10 of which intend to expand them over the coming years. (Facial Recognition Technology: Federal Law Enforcement Agencies Should Better Assess Privacy and Other Risks, United States Government Accountability Office. (Jun. 3, 2021), available at https://www.gao.gov/products/gao-21-518.) One study found that one in four law enforcement agencies across the country can access some form of FRT, and that half of American adults–more than 117 million people–are in a law enforcement face recognition network. (Garvie et al., The Perpetual Line-Up: Unregulated Police Face Recognition in America, Georgetown Law Center on Privacy and Technology (Oct. 18, 2016), available at https://www.perpetuallineup.org/.) Very few of these agencies have a formal facial recognition policy, but one such agency, the New York Police Department, defines the scope of its policy as follows: “Facial recognition technology enhances the ability to investigate criminal activity and increases public safety. The facial recognition process does not by itself establish probable cause to arrest or obtain a search warrant, but it may generate investigative leads through a combination of automated biometric comparisons and human analysis.” (Facial Recognition Technology Patrol Guide, City of New York Police Department (Mar. 12, 2020), available at https://www1.nyc.gov/assets/nypd/downloads/pdf/nypd-facial-recognition-patrol-guide.pdf.)
Proponents of facial recognition technology see it as a useful tool in helping identify criminals. It was reportedly utilized to identify the man charged in the deadly shooting at The Capital Gazette’s newsroom in Annapolis, Maryland in 2018. (Singer, Amazon’s Facial Recognition Wrongly Identifies 28 Lawmakers, A.C.L.U. Says, New York Times, (Jul. 26, 2018), available at https://www.nytimes.com/2018/07/26/technology/amazon-aclu-facial-recognition-congress.html?login=facebook.)
The inaccuracy, biases, and potential privacy intrusions inherent in many facial recognition systems used by law enforcement have led to criticism from civil rights advocates, especially in California. In March 2020, the ACLU, on behalf of a group of California residents, filed a class action lawsuit against Clearview AI, claiming that the company illegally collected biometric data from social media and other websites, and applied facial recognition software to the databases for sale to law enforcement and other companies. (Clearview AI class-action may further test CCPA’s private right of action, JD Supra (Mar. 12, 2020), available at https://www.jdsupra.com/legalnews/clearview-ai-class-action-may-further-14597/.) An investigation by Buzzfeed in 2021 found that 140 state and local law enforcement agencies in California had used or tried Clearview AI’s system. (Your Local Police Department Might Have Used This Facial Recognition Tool To Surveil You. Find Out Here. Buzzfeed News (Apr. 6, 2021). available at https://www.buzzfeednews.com/article/ryanmac/facial-recognition-local-police-clearview-ai-table.)
The controversy surrounding law enforcement use of facial recognition has led many California cities to ban the technology, including San Francisco, Oakland, Berkeley, Santa Cruz and Alameda. Despite the ban in San Francisco, officers there may have skirted the city’s ban by outsourcing an FRT search to another law enforcement agency. (Cassidy, Facial recognition tech used to build SFPD gun case, despite city ban, San Francisco Chronicle (Sept. 24, 2020), available at https://www.sfchronicle.com/bayarea/article/Facial-recognition-tech-used-to-build-SFPD-gun-15595796.php.)
In September 2021, the Los Angeles Times reported that the Los Angeles Police Department had used facial recognition software nearly 30,000 times since 2009, despite years of “vague and contradictory information” from the department “about how and whether it uses the technology.” According to the Times, “The LAPD has consistently denied having records related to facial recognition, and at times denied using the technology at all.” Responding to the report, the LAPD claimed that the denials were just mistakes, and that it was no secret that the department used such technology. Although the department could not determine how many leads from the system developed into arrests, it asserted that “the technology helped identify suspects in gang crimes where witnesses were too fearful to come forward and in crimes where no witnesses existed.” (Despite past denials, LAPD has used facial recognition software 30,000 times in last decade, records show, Los Angeles Times, (Sept. 21, 2020) available at https://www.latimes.com/california/story/2020-09-21/lapd-controversial-facial-recognition-software.)
As noted in the NYPD policy and in the guidelines provided by the developers of the technology, FRT is not supposed to be used as the sole basis for arresting someone. On the contrary, the results it produces instead are intended to assist in an investigation and require taking additional investigative steps According to a recent New York Times investigation:
Law enforcement officers generally say they do not need to mention the use of facial recognition technology because it is only a lead in a case and not the sole reason for someone’s arrest, protecting it from exposure as if it were a confidential informant. But according to Clare Garvie, an expert on the police use of facial recognition, there are four other publicly known cases [beyond the case discussed in the article] of wrongful arrests that appear to have involved little investigation beyond a face match, all involving Black men. She has come across a handful of other examples across the country, she said, in her work with the National Association of Criminal Defense Lawyers. (Hill and Mac, ‘Thousands of Dollars for Something I Didn’t Do,’ New York Times (Mar. 30, 2023), available at https://www.nytimes.com/2023/03/31/technology/facial-recognition-false-arrests.html.)
In another New York Times’ article related to the first known false arrest of a Black man based only on the use of faulty FRT, the facial recognition results explicitly instructed, in all bolded capital letters, “THIS DOCUMENT IS NOT A POSITIVE IDENTIFICATION. IT IS AN INVESTIGATIVE LEAD ONLY AND IS NOT PROBABLE CAUSE TO ARREST. FURTHER INVESTIGATION IS NEEDED TO DEVELOP PROBABLE CAUSE TO ARREST.” (Hill, Wrongfully Accused by an Algorithm, New York Times (Aug. 3, 2020) available at https://www.nytimes.com/2020/06/24/technology/facial-recognition-arrest.html). That man, Robert Williams, was arrested and held in jail, apparently solely on the bases of the FRT results, for a burglary at a store he had not been in since 2014 and that he had an alibi for. (For a detailed account of his experience, please see ARGUMENTS IN OPPOSITION )
4) Author’s statement. According to the author:
I authored AB 1215 in 2019 which banned the use of biometric surveillance through police body cameras. The bill only passed with a three year moratorium that expired January 1, 2023. Consequently, current law has absolutely no parameters or restrictions regarding law enforcement’s use of facial recognition technology. It is critical that we ensure there are safeguards in place in order to avoid another year of unregulated use. We can’t go another year with no protections. AB 642 is a response to a battle that we cannot afford to risk losing. The bill includes critical safeguards such as codifying an accuracy level, accountability and oversight, reporting, civil penalties, protections for immigrant communities, people seeking services such as abortion and gender affirming care, prohibits law enforcement from using a match alone to arrest someone, to request a warrant, to violate someone’s constitutional rights, and to discriminate against protected characteristics. Most importantly, this bill does not prohibit nor deter local governments from choosing to ban the use of facial recognition technology.
5) How this bill would regulate the use of facial recognition technology by law enforcement. According to the author, in an effort to strictly regulate the use of FRT, the bill is intended to work as follows:
1. Defines facial recognition technology as, in part, including any surveillance system that actively uses FRT to identify people in a surveilled area in real time.
2. Allows the broad use of virtually all available data bases, including arrest photograph databases and reference photograph databases owned by third parties.
3. Establishes the authorized uses for FRT, including identifying a person the officer reasonably believes is suspected of committing a felony; identifying a dead or missing person; identifying a person who has been arrested; or identifying any person if the officer decides an emergency situation exists.
4. Establishes prohibited uses of FRT, including identifying people exercising First Amendment rights, people seeking gender affirming care or abortion services, or immigrants.
4. Establishes oversight provisions, including the information that is required to be documented when the technology is used; preparing and submitting an annual report to the California State Auditor; and requiring the State Auditor, in turn, to produce an annual report.
5. Requires the arrest database being used to be purged of certain images every six months.
6. Establishes acceptable accuracy standards, which are a 98 percent accuracy rating from the National Institute of Standards and Technology (NIST) and prohibits an officer from using the technology in a way that reduces the accuracy rating.
7. Requires when FRT is used to identify a person who is subsequently arrested, that the person must be provided with certain information.
8. Establishes civil penalties for intentionally violating the statute.
9. Sets a 2-year statute of limitations for civil suits.
6) Committee amendments. The Committee amendments are intended as a modest first step towards addressing the privacy issues that are of concern in this bill. There are four changes:
1. Clarification that the prohibitions in the bill apply not just to individual officers, but to the entire agency, whether the person using the database for a prohibited use is a sworn police officer or not.
A peace officer using or requesting the use of FRT shall not do any of the following:
Will now be --
A law enforcement agency or officer authorizing the use of, maintaining the database for, or using FRT shall not do any of the following:
2. Tightening of the language in the bill designed to protect people seeking abortion services and gender affirming care to mirror other laws prohibiting the sharing of data related to these services.
(d) (1) Provide the results of, or information derived from, the use of FRT to any individual or to any agency or department in another state regarding the provision of lawful gender-affirming health care or gender-affirming mental health care performed in this state.
Will now be --
(d)(1) Notwithstanding any other provision of this title, a law enforcement agency or officer shall not release any FRT data or results, or allow access to databases used with FRT, in response to a subpoena or request if that subpoena or request is based on any of the following:
- Another state’s laws that interfere with a person’s rights under the Reproductive Privacy Act (Article 2.5 (commencing with Section 123460) of Chapter 2 of Part 2 of Division 106 of the Health and Safety Code.
- A foreign penal civil action, as defined in Section 2029.200 of the Code of Civil Procedure.
- An investigation related to a natural person seeking gender affirming health care or gender affirming mental health care.
3. Similar to #2, tightening the language in the bill which prohibits the sharing of data with federal agencies seeking to enforce federal immigration law.
(c) Share FRT data with any state or federal agency for the purpose of enforcing federal immigration law.
Will now be --
(c) Notwithstanding any other provision of this title, a law enforcement agency or officer shall not release any FRT data or results, or allow access to the databases used with FRT, in response to a subpoena or request if that subpoena or request is in violation of the California Values Act.
4. Adding a requirement that the Department of Technology, the state’s primary repository of cybersecurity expertise, issue standards related to FRT data.
(b) (1) By July 1, 2024, the California Department of Technology, in consultation with the Chief of the Office of Information Security, shall issue standards to ensure the confidentiality and cybersecurity of FRT data and results. A law enforcement agency is required to establish a written policy that adheres to these standards. Nothing in this paragraph shall prevent a law enforcement agency from adopting a written policy that provides for standards stronger than those issued by the Department.
(2) A law enforcement agency that uses FRT shall have a written policy that includes, without limitation, all of the following:
(3) A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT by the Commission on Peace Officer Standards and Training.
(4) A requirement that a manager authorized to use FRT be assigned to oversee the FRT program.
(5) A policy that describes the parameters of acceptable inputs to be used as probe images and that prohibits the use of sketches or other manually produced images.
(6) An acceptable use policy that includes specific allowances and restrictions on use.
7) Significant issues remain with this bill. As noted in the previous section on Committee amendments, the changes made to the bill in this Committee are not meant to be all encompassing, nor do they resolve the largest concerns with the bill. The author has stated that he is attempting to establish real and robust guardrails limiting the use of FRT, that these are complicated issues and would benefit from the input of a broad range of experts and stakeholders. The expectation is that if this bill passes out of the Committee, the larger issues will be addressed prior to it coming back to the Assembly for concurrence. The following is a list of the major issues identified by the Committee:
1. The bill contains broad permission to use any FRT system on images from any device, including real-time identification of people from body-worn camera video, against any database chosen. This latitude may be excessive.
A. As currently written, this bill allows for the adoption of any type of FRT (see 1) a) in the SUMMARY). While perhaps intended to clarify that any future technology will fall under this statute, this broad definition appears to give tacit approval to law enforcement agencies to use any FRT system, regardless of its privacy implications. Robust guardrails ought to precisely enumerate how and what technology can be used and prohibit other uses without express permission from the Legislature.
B. As currently written, the bill allows law enforcement to use probe images, e.g., an image from a closed circuit video, body-worn camera, dash camera, or any other surveillance camera, on virtually any database. Rather than being limited to running the photo through their mug shot database only, the FRT vendor can use any database, whether government-developed or one owned by a third party, e.g., Facebook, to search through hundreds of millions, if not billions, of images. This allows law enforcement to freely search the images of millions of people without first obtaining a court order, warrant, or some other permission to do so.
2. The 98% NIST accuracy requirement likely will not improve the accuracy of the results. A 98 percent NIST rating does not appear to mean the results are accurate 98 percent of the time. It means that during the controlled NIST testing, 98 percent of the time, the system accurately identified the person in the picture along with an unknown number of other people that the system also identified as possibly being the person. In other words, the correct person could be among one of 50 other people who were also identified as possibly being the person.
In addition, as discussed previously, an overall 98 percent accuracy rating, is not the same as determining that the system was able to identify the person in the photo 98 percent of the time when tested against all demographics.
Finally, the NIST test determines the accuracy rating using a particular database and particular types of photographs, e.g., an employee identification photo compared to a database containing all of the employees of a company. It does not test the ability of the system to not identify potential matches when the person in the probe photo is not in the database, for example. Also, it is important to note, that in order to guarantee a 98 percent accuracy rating, NIST would need to test every probe photo on the specific database being used to determine the accuracy level.
3. Significant bias remains in FRT systems, making their use dangerous. While this issue was discussed in detail previously, it is worth repeating. NIST’s 98% accuracy score does not account for racial bias. When looking at one algorithm that received a 99% accuracy rating from NIST, the false positive identification rate (FPIR) for Black men was more than 2x the FPR for white men, and for a couple of the thresholds, the disparity was more than 3x. Thus, NIST’s own testing results indicate that an algorithm that may clear the standard in the bill may have a false positive rate for Black men 3x the false positive rate for white men. (Those test results that bear this out are available at https://pages.nist.gov/frvt/reports/demographics/annexes/annex_16.pdf. )
In a 2020 study on facial recognition in body worn cameras, “the researcher conducted the study in conditions that were generally stable and controllable, yet matching performance error rates were as high as 100%." Notably, this conclusion is tied to two aspects of body cameras that aren't likely to change: the footage is the result of officers moving, and the footage is filmed with a wide angle, which skews faces. Importantly, as noted, the study was done in conditions that are unlikely to be the conditions officers encounter in the field where people are continually moving – including the officer.
4. The bill provides broad latitude for an officer to use the system in ways that are intended to be prohibited. Specifically, bill language currently includes words and phrases that dilute the guidelines. As an example, section 5 in the SUMMARY describes authorized situations in which FRT can be used (e.g., to identify someone suspected of committing a felony or someone who is missing or dead). While these limitations may seem clear and sensible, 5) d) in that section waters down those prohibitions by allowing an officer to use the system if the officer determines an emergency situation exists that involves immediate danger of death or serious injury.
In addition, the specific prohibition against using FRT on people who are exercising their First Amendment rights states that officers cannot use the system to “identify any persons solely on the basis that the person is exercising their rights.” (10) a) in the SUMMARY.)
Using the exceptions in both of these parts of the bill, for example, means that FRT could in fact be used by an officer who is doing crowd control at a protest where people are exercising their First Amendment rights. In order to freely surveil the protesters, all the officer would need to assert is that they were concerned that the protest may devolve and someone may get injured or killed therefore it was a permissible use.
5. The bill lacks robust oversight and data collection requirements. Under the bill, the State Auditor is tasked with collecting information and overseeing the use of the technology. The State Auditor lacks strong enforcement authority and, regardless of the language in this bill, could be tasked with auditing agencies’ use of FRT at any time. In developing an oversight framework that seeks to create robust oversight and data collection, the Legislature should consider tasking the Attorney General with oversight and vest the office with the power to prohibit the use of the technology in the event it is being misused or proper security measures are not in place. In addition, any legislation should include comprehensive data collection that will be uploaded in the OpenJustice data portal allowing for open access to the data for policy-makers, the public and researchers.
6. The bill lacks any cybersecurity requirements and speaks only to the regulation of law enforcement agencies, not restrictions on third-party vendors. Law enforcement agencies that take advantage of this technology are not purchasing a system that is self-contained and housed within their agency. They are contracting with third-party vendors that run the images for them.
As has been demonstrated by the actions of companies like Clearview AI, any legislation that contemplates regulating the use of FRT should consider the relationship between the law enforcement agency and the vendor. At a minimum, the bill should contemplate a prohibition against images shared by California law enforcement being added to the vendors’ databases.
7. Lacks any training requirements for the use of the system. This bill requires that officers receive training from the Commission on Peace Officer Standards and Training (POST) prior to using FRT. However, a review of the POST catalog shows that there is currently no POST certified training on the appropriate use of FRT. This lack of training poses challenges for the bill as it is currently drafted. In considering a regulatory framework more in-depth discussion related to what the training must include and how often it needs to be taken should be considered.
Beyond the specific concerns of the bill, however, the lack of POST certified training means that many law enforcement agencies around the state are using this technology without any proper training and may not understand that the results are not definitive and can be further compromised by poor quality probe images.
8. Regardless of the restrictions put in place in state law, experience from departments around the country and recent discoveries related to the use of Automated License Plate Reader (ALPR) data in California suggests that they may be ignored, putting people in danger. In contemplating the regulation of FRT versus restricting or prohibiting its use, the Legislature should use the experience with ALPR systems as a guide, particularly the findings of the State Auditor’s office as a result of their 2019 audit. (See this Committee’s analysis of AB 1436 (Lowenthal, 2023) for a detailed discussion of the report.)
Experience in California and around the country has shown that not infrequently, law enforcement officers and agencies are either unaware of or choose to ignore state laws designed to establish limits and restrictions related to policing. This is already being seen in the misuse of FRT. According to a recent New York Times investigation:
Law enforcement officers generally say they do not need to mention the use of facial recognition technology because it is only a lead in a case and not the sole reason for someone’s arrest, protecting it from exposure as if it were a confidential informant. But according to Clare Garvie, an expert on the police use of facial recognition, there are four other publicly known cases [beyond the case discussed in the article] of wrongful arrests that appear to have involved little investigation beyond a face match, all involving Black men. She has come across a handful of other examples across the country, she said, in her work with the National Association of Criminal Defense Lawyers. (Hill and Mac, ‘Thousands of Dollars for Something I Didn’t Do,’ New York Times (Mar. 30, 2023), available at https://www.nytimes.com/2023/03/31/technology/facial-recognition-false-arrests.html.)
[A letter in opposition to this bill from one of the five men identified by Clare Garvie, Robert Williams, is included in its entirety at the end of this analysis.]
8) Final analysis. The question before this Committee is whether or not this bill furthers the Committee’s policy priorities. First and foremost, protecting Californians’ constitutional right to privacy. Along with that, the Committee is working to ensure that all Californians, and those coming from out of state, are protected from punitive and discriminatory draconian laws attacking the LGBTQ+ community and criminalizing people seeking abortion and gender affirming care. A further priority of the Committee is ensuring that the State’s laws protect our immigrant neighbors from federal policies that make them vulnerable to being separated from their families, imprisoned, and ultimately returned to countries that many were often forced to flee from for their own safety. The answer to this question when it comes to this bill is that in its current version it does not, but with further amendment, it is hoped that it will.
As the author notes, law enforcement agencies are currently using FRT around the state without restriction or regulation. Given the faulty nature of the technology, Californians would likely be well served by robust regulation and strict limits on its use.
However, the larger question before the Legislature this year remains, has the technology reached a stage where it can be used in a restricted manner to assist in law enforcement investigations? Based on the current research discussed previously, this is an open question.
9) Related legislation. AB 1034 (Wilson, 2023) would prohibit a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other officer camera. AB 1034 is being heard in this Committee today.
AB 793 (Bonta, 2023) would provide that a government entity may not seek, from any court, a compulsory process to enforce a reverse-location demand or a reverse-keyword demand, as defined. AB 793 is pending hearing in the Assembly Appropriations Committee.
AB 1483 (Lowenthal, 2023) requires a local public agency end-user of an automated license plate reader (ALPR) to purge information that does not match information on a hot list, as defined, within 30 days and explicitly prohibits the selling, sharing or transferring of ALPR data with an out-of-state or federal agency without a valid California court order or warrant. That bill is currently pending before this Committee.
AB 1281 (Chau, 2019) would have required any business that uses facial recognition technology in California to disclose that usage in a physical sign that is clear and conspicuous at the entrance of every location. This bill was placed on the inactive file on the Senate Floor.
AB 1215 (Ting, Ch. 579, Stats. 2019) prohibits a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other camera. It sunsetted effective January 1, 2023.
AB 375 (Chau, Ch. 55, Stats. 2018) enacted the CCPA to ensure the privacy of Californians’ personal information through various consumer rights.
SB 1121 (Dodd, Ch. 735, Stats. 2018) ensured that a private right of action applied only to the CCPA’s section on data breach and not to any other section of the CCPA, as specified, corrected numerous drafting errors, made non-controversial clarifying amendments, and addressed several policy suggestions made by the AG in a preliminary clean-up bill after the passage of AB 375.
10) Next steps for this bill. Should this bill pass, the author has committed to working on the issues raised in this Committee analysis. It is expected the bill will be amended further to address these issues. As an example, it may be appropriate, for example, to prohibit the use of any facial recognition technology until NIST determines that all biases have been eliminated. Discussions around this bill’s provisions are far from over, and if the bill is not amended to satisfy the concerns of the Committee, it is likely that this Committee will assert its power to re-hear the bill on concurrence.
ARGUMENTS IN SUPPORT: 18 law enforcement entities that were in support of the previous version of the bill, have removed their support. The League of California Cities remains in support and writes:
The League of California Cities (Cal Cities) is pleased to support AB 642 (Ting). This measure would require any law enforcement agency that uses facial recognition technology (FRT) to have a written policy governing the use of that technology and would require any FRT system used to meet certain national standards and would limit the use of FRT to use as an investigative aid. Additionally, the measure would specifically prohibit the use of any FRT-generated match from being the sole basis for probable cause in an arrest, search, or warrant and would also require an agency using FRT to post their written policy and an annual summary of FRT usage on their internet website.
Facial recognition technology is one of many tools utilized in identifying an individual by comparing a digital image of the person’s face to a database of known faces, typically by measuring distinct facial features and characteristics. This technology does not by itself result in ultimate identification, but it may generate investigative leads necessary for combatting crime within our communities. Technology assists our law enforcement partners in doing their jobs more efficiently and ultimately improves public safety.
Cal Cities supports accountability on the part of law enforcement agencies concerning police technology and policies, as well as related oversight by local governing bodies. However, we do not support policies that restrict law enforcement agencies from utilizing technologies that would otherwise enhance their ability to prevent criminal activity in the communities they serve.
ARGUMENTS IN OPPOSITION: Over 50 civil rights and social justice organizations are opposing this bill. In addition, Robert Williams, the first man who has been identified as being arrested because of faulty FRT, has also written in opposition to this bill. His letter is included here in its entirety so that it can be included in the record.
Statement of Robert Williams.
Three years ago, I was arrested for a crime I didn’t commit based on a false face recognition match. The nightmarish ordeal upended my life and convinced me that law enforcement should not have access to this inherently flawed, racially biased technology.
My story began in January 2020 with a phone call from Detroit police advising me to turn myself in. They refused to tell me why, so I assumed it was a prank.
When I arrived home, police were waiting for me. With no explanation, officers handcuffed me on my front lawn while my distressed wife and young daughters watched. At the detention center, officers took my fingerprints, DNA sample and mugshot. Scared and confused, I spent the night on the cold and filthy concrete floor of an overcrowded cell.
The next day, police accused me of stealing thousands of dollars’ worth of watches from a store I hadn’t visited in years.
I could prove I was driving home from my job 40 minutes outside Detroit at the time of the robbery. It didn’t matter. Facial recognition software had matched a blurry image pulled from the store’s grainy surveillance footage to my driver’s license photo, and a security consultant who watched the video, but who had not seen the suspect in person, picked me out of a shoddy photo lineup. Even though police were specifically told the facial recognition result was just a lead and was not to be used as the sole basis for an arrest, for the detectives, it was enough to arrest me.
Even after police acknowledged the face recognition software had made a mistake, they didn’t immediately release me. I spent 30 hours in custody before I was allowed to return to my worried family. After dropping the charges, in August 2020 the Wayne County prosecutor’s office expunged my record and deleted my fingerprints from the police department’s database.
I may be the first documented case of a wrongful arrest based on a false face recognition match, but I wasn’t the last. In the years following my arrest, police have similarly misidentified and detained four other individuals we know of across the country. All of us are Black men. All of us suffered the trauma of being wrenched from our families, isolated, and interrogated by police officers who dismissed our claims of innocence because they believed the technology was infallible. In fact, studies have found that Black and Asian people are up to 100 times more likely to be misidentified than white men.
In my case, Detroit police were supposed to treat face recognition matches as an investigative lead, not as the only proof they need to charge someone with a crime. They should have collected corroborating evidence such as an eyewitness identification, cell phone location data or a fingerprint. They had none of that – just an out-of-focus image of a large Black man in a baseball cap that a faulty algorithm had determined was me.
My message to this committee is that AB 642 is dangerous. It would be a grave mistake to assume that AB 642 will prevent what happened to me from happening to people in California. As in my situation, this bill instructs officers not to solely rely on facial recognition results.
Regardless of the bill’s language, if the software identifies a match, police will think they’ve found the right person. Once that happens, they will zero in on that individual, subjecting them to unjustified scrutiny, disregarding evidence that doesn’t fit their chosen narrative, and ignoring other potential suspects. Face recognition matches would become a driving force behind police investigations and false positives will lead to bad arrests.
There is no acceptable number of misidentifications. One out of 50, one out of 100, one out of 10,000, it’s all too high. No one should have to go through what I did, but I fear if this bill is passed, they will.
If the California legislature embraces the widespread use of face recognition technology, there will be many more cases like mine – innocent people put in harm’s way and damage that can never be undone. But they may not be as lucky as I was to avoid a conviction, or worse, a fatal encounter with police.
False matches don’t only harm the person who is arrested. The experience traumatizes the entire family. My daughters have not forgotten seeing their father shoved into the back of a patrol car. To this day, the frightening memory sometimes makes them cry.
Face recognition software is flawed and dangerous. None of the provisions in this bill will adequately prevent police from abusing the technology and arresting someone simply because a racially biased algorithm scanned their face. The only safe path forward is a total prohibition on police use of face recognition.
REGISTERED SUPPORT / OPPOSITION:
Support
League of California Cities
Orange County Board of Supervisors - Supervisor Vicente Sarmiento
Opposition
A New Way of Life
Access Reproductive Justice
ACLU California Action
All of Us or None - Riverside
Anti Police-Terror Project
Asian Americans Advancing Justice - Asian Law Caucus
Asian Law Alliance
Bend the Arc California
Bend the Arc: Jewish Action
California Association of Black Lawyers
California Coalition for Women Prisoners
California Immigrant Policy Center
California Innocence Coalition: Northern California Innocence Project, California Innocence Project, Loyola Project for The Innocent
California United for A Responsible Budget (CURB)
Cancel the Contract
Care First California
Central American Resource Center of San Francisco
Citizens for A Better Los Angeles
Clergy and Laity United for Economic Justice
Coalition for Homelessness San Francisco
Council on American Islamic Relations
Electronic Frontier Foundation
Equal Justice Society
Family Reunification Equity & Empowerment (F.R.E.E.)
Fight for The Future
Free Speech Coalition
If/when/how: Lawyering for Reproductive Justice
Indivisible CA Statestrong
Initiate Justice
Lawyers' Committee for Civil Rights of The San Francisco Bay Area
Legal Services for Prisoners With Children
Long Beach Immigrant Rights Coalition
Media Alliance
MediaJustice
Muslim Democrats and Friends
National Center for Lesbian Rights
Oakland Privacy
Orange County Rapid Response Network
Partnership for The Advancement of New Americans
People's Budget Orange County
Positive Women's Network - USA
Privacy Rights Clearinghouse
Safer Streets LA
San Francisco Public Defender - Racial Justice Committee
San Jose Nikkei Resisters
Secure Justice
Silicon Valley De-bug
St James Infirmary
Starting Over INC.
Stop the Musick Coalition
Tenth Amendment Center
Training in Early Abortion for Comprehensive Healthcare (TEACH)
Transforming Justice Orange County
Transgender, Gendervariant, Intersex Justice Project
Urge: Unite for Reproductive & Gender Equity
2 Individuals
Analysis Prepared by: Julie Salley / P. & C.P. / (916) 319-2200
Date of Hearing: April 11, 2023
Counsel: Andrew Ironside
ASSEMBLY COMMITTEE ON
Reginald Byron Jones-Sawyer, Sr., Chair
ABPCA Bill Id:
Author:() – As Ver:
SUMMARY:
Sets minimum standards for use of facial recognition technology (FRT) by law enforcement, including requiring law enforcement agencies to have a written policy for FRT use, allowing for FRT use when a peace officer has reasonable suspicion that an individual has committed a felony, and providing that an FRT-generated match of an individual may not be the sole basis for probable cause for an arrest, search, or affidavit for a warrant. Specifically, this bill:
- Defines “facial recognition technology” or “FRT” as a system that compares a probe image of an unidentified human face against a reference photograph database, and, based on biometric data, generates possible matches to aid in identifying the person in the probe image.
- Provides that FRT includes any surveillance system that actively uses FRT to identify persons in a surveilled area in real time.
- Authorizes a peace officer to use or request the use of FRT for any of the following reasons:
- To assist in identifying a person that officer has reasonable suspicion to believe has committed a felony;
- To assist in identifying a person who is deceased or who has been reported missing, as specified;
- To assist in identifying any person who has been lawfully arrested, during the process of booking or during that person’s custodial detention; or,
- To assist in identifying any person if a peace officer determines that an emergency situation exists that involves immediate danger of death or serious physical injury to any person and the identification of that person is necessary to prevent that death or injury.
- To assist in identifying a person that officer has reasonable suspicion to believe has committed a felony;
- Requires, if a peace of uses or requests the use of FRT, to document all of the following information:
- The identity of the peace officer using or requesting the use of FRT, and, if applicable, the officer authorizing the use or request;
- A detailed description, as available, of the person being identified;
- Any photograph or video being used as a probe image; and,
- Any details regarding other investigative measures taken to identify the person and an explanation of why those measures failed or are reasonably unlikely to succeed.
- The identity of the peace officer using or requesting the use of FRT, and, if applicable, the officer authorizing the use or request;
- Requires the custodian of any arrest photograph database being used in conjunction with FRT, beginning on July 1, 2024 and every six months thereafter, to remove from the database any of the following images:
- Any photograph of a person under 18 years of age;
- Any photograph of an arrested or detained person who has been released without being charged with an offense, or whose charges have been dismissed;
- Any photograph of an arrested or detained person who has subsequently been acquitted of the charged offense; or,
- Any photograph of a person whose conviction has been expunged, has been exonerated of the crime, or who has had their conviction reversed on appeal.
- Any photograph of a person under 18 years of age;
- Provides that the image removal requirement applies only to the use of a reference photograph database for the use of FRT and shall not be construed to prohibit a peace officer from using any other investigative database including a fingerprint database.
- Requires any agency that maintains and operates an arrest photograph database to establish procedures to ensure compliance with the image removal requirement.
- Prohibits a peace officer using or requesting the use of FRT from doing any of the following:
- Using FRT to identify any person solely on the basis that the person is exercising rights guaranteed by the United States Constitution, including free assembly, association, and speech;
- Relying on actual or perceived race, ethnicity, national origin, religion, disability, gender, gender identity, or sexual orientation in selecting a person to identify using FRT, except when there is trustworthy information, relevant to the locality and timeframe, in the context of a particular area and for a particular period of time, that links a person with a particular characteristic described to an identified criminal incident or scheme;
- Sharing FRT data with any state or federal agency for the purpose of enforcing federal immigration law;
- Providing the results of, or information derived from, the use of FRT to any individual or to any agency or department in another state regarding the provision of lawful gender-affirming health care or gender-affirming mental health care performed in this state;
- Providing the results of, or information derived from the use of FRT to any individual or agency or department in another state regarding the provision of abortion services in this state;
- Using an FRT match as the sole basis upon which probable cause is established for a search, arrest, or affidavit for a warrant, and provides that any peace officer using information obtained from the use of FRT shall examine results with care and consider the possibility that matches could be inaccurate; and,
- Using FRT in conjunction with any reference photograph databse that contains information, including images, obtained by any of the following means:
- In a manner that violates federal or state law;
- In a manner that violates a service agreement between a provider of an electronic communication service to the public or a provider of a remote computing service and customers or subscribers of that provider;
- In a manner that is inconsistent with the privacy policy of a provider, as specified;
- By deceiving a person whose information was obtained;
- Through the unauthorized access of an electronic device or online account;
- In violation of a contract, court settlement, or other binding legal agreement; or,
- From unlawful or unconstitutional practices by any governmental official or entity.
- In a manner that violates federal or state law;
- Using FRT to identify any person solely on the basis that the person is exercising rights guaranteed by the United States Constitution, including free assembly, association, and speech;
- Provides that, for the purpose of prohibiting sharing FRT results regarding persons receiving abortion servicers or gender-affirming health care and gender-affirming mental health care, “information derived from the use of FRT” means information that would not have been discovered or obtained but for the use of FRT, regardless of any claim that the information would inevitably have been discovered or obtained through other means.
- Requires each law enforcement agency using FRT or requesting the use of FRT of another agency, by no later than January 31, 2025, and annually thereafter, to prepare and submit a report to the California State Auditor containing only the following information regarding the use of FRT, as applicable:
- The information a peace officer is required to document when using or requesting the use of FRT;
- Whether modifications were made to any probe images and what those modifications were;
- The arrests that FRT results contributed to, and the offenses for which the arrests were made, disaggregated by race, ethnicity, gender, and age;
- A description of the reference photograph database that was used;
- A description of the FRT system that was used;
- The total number of searches performed;
- The total number of times FRT results were shared with another law enforcement agency, an outside agency, or a third party; and,
- The actions taken to comply with requirement that a law enforcement agency remove certain images from a reference photograph database, as specified.
- The information a peace officer is required to document when using or requesting the use of FRT;
- Requires each district attorney’s office, city prosecutor’s office, and the Attorney General, by no later than January 31, 2025 and annually thereafter, to report to the California State Auditor the following information regarding information obtained from FRT:
- The number of convictions that FRT results contributed to and the offenses for which the convictions were obtained, disaggregated by race, ethnicity, gender, and age; and
- The number of motion to suppress made related to FRT results, and the number granted or denied.
- The number of convictions that FRT results contributed to and the offenses for which the convictions were obtained, disaggregated by race, ethnicity, gender, and age; and
- Requires the State Auditor, on or before July 1 of each year, to release to the public, post online, and transmit to the Legislature a full and complete report concerning the use of FRT by law enforcement agencies and prosecutors, including any violations identified.
- Requires any law enforcement agency using FRT to keep and maintain FRT activity logs or other required records, as specified.
- Prohibits a law enforcement agency from operating an FRT system that has not been evaluated under the National Institute of Standards and Technology Face recognition Vendor Testing Program and achieved an accuracy score of 98 percent true positives within two or more datasets relevant to investigative applications on a program report.
- Prohibits a peace officer from using FRT in any manner that reduces the program’s competency score below 98 percent.
- Requires a law enforcement agency that uses FRT to have a written policy that includes, without limitation, all of the following:
- A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT by the Commission of Peace Officer Standards and Training;
- A requirement that a manger authorized to use FRT be assigned to oversee the FRT program;
- A policy that describes the parameters of acceptable inputs to be used as probe images and that prohibits the use of sketches or other manually produced images; and,
- An acceptable use policy that includes specific allowances and restrictions on use.
- A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT by the Commission of Peace Officer Standards and Training;
- Requires each law enforcement agency using FRT, by no later than July 1, 2024, to post the required written policy on its website.
- Prohibits a law enforcement agency or peace officer from requesting or entering into an agreement with another law enforcement agency or other third party to perform FRT search on behalf of the requesting officer or agency if the program operated by the other agency or party does not meet the specified accuracy requirements.
- Requires a law enforcement agency, if the State Auditor identifies any violations by that agency, to cease using FRT until all violations have been corrected.
- Requires the law enforcement agency to notify the public if its use of FRT is suspended for violations, as specified.
- Requires a law enforcement agency that uses FRT to attempt to identify an individual who is arrested to provide the individual with both of the following:
- A notice of the name of the law enforcement agency that operated the FRT system used, and the name of the database, if any, the was used to identify the individual; and
- A copy of the required accuracy or bias report, each probe image that was used by the agency, any modifications made to the probe image, the candidate list, in rank order, produced by the facial recognition system, and any other documentation related to the use of FRT in the law investigation.
- A notice of the name of the law enforcement agency that operated the FRT system used, and the name of the database, if any, the was used to identify the individual; and
- Requires the notice provided by the law enforcement agency to the individual arrested to be an appropriate language for the person if they are not fluent or literate in English.
- Provides that, if a court or law enforcement agency determines that a peace officer has used FRT in violation of the law, and the court or agency finds that the circumstances surrounding the violation raise serious questions about whether or not the officer acted intentionally with respect to the violation, the agency shall promptly initiate a proceeding to determine whether disciplinary action against the officer is warranted.
- Provides that, notwithstanding any other law, a violation of this title is not punishable as a crime.
- Authorizes a person who is subject to identification or attempted identification through FRT in violation of the law to bring a civil action against the peace officer or law enforcement agency responsible for the violation.
- Provides that a person who is the subject of disparate treatment or adverse impact on the basis of race, ethnicity, gender, or age, whether individually or as a member of a class of individuals, due to use of FRT or any technological element, criteria, method, or design feature thereof, by a law enforcement agency, may bring a cause of action against the peace officer, law enforcement agency, or maker of the facial recognition or face surveillance technology responsible for the violation.
- Provides that the following relief may be recovered or obtained in a civil action for an FRT use violation:
- Any actual damages sustained by that person as a result of the violation;
- The greater of either of the following:
- Statutory damages of $50,000 per violation; or
- Profits earned as a result of each violation.
- Statutory damages of $50,000 per violation; or
- Exemplary damages, as specified;
- Injunctive, equitable, or declaratory relief as may be appropriate, including preliminary injunctive relief; and,
- Costs of the cation, together with reasonable attorney’s fees.
- Any actual damages sustained by that person as a result of the violation;
- Provides that a good faith reliance on a statutory authorization, or a good faith determination that the conduct complained of is permitted, is a complete defense against a civil action.
- Prohibits a civil action from commencing later than two years after the date upon which the claimant discovered or first had a reasonable opportunity to discover the violation.
- Defines “arrest photograph database” as a database populated primarily by booking or arrest photographs or other photographs of persons with law enforcement contacts.
- Defines “probe image” as an image of a person that is searched against a database of known, identified persons or an unsolved photograph file.
- Defines “reference photograph database” as a database populated with photographs of individuals that have been identified, including databases composed of driver’s licenses or other documents made or issued by or under the authority of the state, a political subdivision thereof, databases operated by third parties, and arrest photograph databases.
- Provides that FRT does not include any access control system used by a law enforcement agency that uses biometric inputs to confirm the identity of employees or other approved persons for the purpose of controlling access to any secured place, device, or system.
EXISTING LAW:
- Declares that it is the intent of the Legislature to establish policies and procedures to address issues related to the downloading and storage data recorded by a body-worn camera worn by a peace officer; these policies and procedures shall be based on best practices. (Pen. Code, § 832.18, subd. (a).)
- Encourages agencies to consider best practices in establishing when data should be downloaded to ensure the data is entered into the system in a timely manner, the cameras are properly maintained and ready for the next use, and for purposes of tagging and categorizing the data. (Pen. Code, § 832.18, subd. (b).)
- Encourages agencies to consider best practices in establishing specific measures to prevent data tampering, deleting, and copying, including prohibiting the unauthorized use, duplication, or distribution of body-worn camera data. (Pen. Code, § 832.18, subd. (b)(3).)
- Encourages agencies to consider best practices in establishing the length of time that recorded data is to be stored. States that nonevidentiary data including video and audio recorded by a body-worn camera should be retained for a minimum of 60 days, after which it may be erased, destroyed, or recycled. Provides that an agency may keep data for more than 60 days to have it available in case of a civilian complaint and to preserve transparency. (Pen. Code, § 832.18, subd. (b)(5)(A).)
- States that evidentiary data including video and audio recorded by a body-worn camera should be retained for a minimum of two years under any of the following circumstances:
- The recording is of an incident involving the use of force by a peace officer or an officer-involved shooting;
- The recording is of an incident that leads to the detention or arrest of an individual; or,
- The recording is relevant to a formal or informal complaint against a law enforcement officer or a law enforcement agency. (Pen. Code, § 832.18, subd. (b)(5)(B).)
- States that the recording should be retained for additional time as required by law for other evidence that may be relevant to a criminal prosecution. (Pen. Code, § 832.18, subd. (b)(5)(C).)
- Instructs law enforcement agencies to work with legal counsel to determine a retention schedule to ensure that storage policies and practices are in compliance with all relevant laws and adequately preserve evidentiary chains of custody. (Pen. Code, § 832.18, subd. (b)(5)(D).)
- Encourages agencies to adopt a policy that records or logs of access and deletion of data from body-worn cameras should be retained permanently. (Pen. Code, § 832.18, subd. (b)(5)(E).)
- Encourages agencies to include in a policy information about where the body-worn camera data will be stored, including, for example, an in-house server which is managed internally, or an online cloud database which is managed by a third-party vendor. (Pen. Code, § 832.18, subd. (b)(6).)
- Instructs a law enforcement agency using a third-party vendor to manage the data storage system, to consider the following factors to protect the security and integrity of the data: Using an experienced and reputable third-party vendor; entering into contracts that govern the vendor relationship and protect the agency’s data; using a system that has a built-in audit trail to prevent data tampering and unauthorized access; using a system that has a reliable method for automatically backing up data for storage; consulting with internal legal counsel to ensure the method of data storage meets legal requirements for chain-of-custody concerns; and using a system that includes technical assistance capabilities. (Pen. Code, § 832.18, subd. (b)(7).)
- Encourages agencies to include in a policy a requirement that all recorded data from body-worn cameras are property of their respective law enforcement agency and shall not be accessed or released for any unauthorized purpose. Encourages a policy that explicitly prohibits agency personnel from accessing recorded data for personal use and from uploading recorded data onto public and social media Internet websites, and include sanctions for violations of this prohibition. (Pen. Code, § 832.18, subd. (b)(8).)
- Requires that a public agency that operates or intends to operate an Automatic License Plate Recognition (ALPR) system to provide an opportunity for public comment at a public meeting of the agency's governing body before implementing the program. (Civil Code, § 1798.90.55.)
- Prohibits a local agency from acquiring cellular communications interception technology unless approved by its legislative body. (Gov. Code, § 53166, subd. (c)(1).)
FISCAL EFFECT:
Unknown
COMMENTS:
- Author's Statement: According to the author, “I authored AB 1215 in 2019 which banned the use of biometric surveillance through police body cameras. The bill only passed with a three year moratorium that expired January 1, 2023. Consequently, current law has absolutely no parameters set regarding law enforcement’s use of facial recognition technology. It is critical that we ensure there are safeguards in place in order to avoid another year of unregulated use. We can’t go another year with no protections. AB 642 is a response to a battle that we cannot afford to risk losing. The bill includes critical safeguards such as codifying an accuracy level, oversight and reporting, and prohibits law enforcement from using a match alone to arrest someone, to request a warrant, to violate someone’s constitutional rights, and to discriminate against protected characteristics. Most importantly, this bill does not prohibit nor deter local governments from choosing to ban the use of facial recognition technology.”
- Facial Recognition Technology: Facial recognition technology is capable of identifying an individual by comparing a digital image of the person’s face to a database of known faces, typically by measuring distinct facial features and characteristics. Early versions of the technology were pioneered in the 1960s and 1970s, but true facial recognition technology as we understand it today did not come about until the early 1990s. In 1993, the United States military developed the Facial Recognition Technology (FERET) program, which aimed to create a database of faces and recognition algorithms to assist in intelligence gathering, security, and law enforcement. (“Facial Recognition Technology (FERET).”) The National Institute of Standards and Technology, United States Department of Commerce. https://www.nist.gov/programs-projects/face-recognition-technology-feret).) Since that time, advances in computer technology and machine learning have led to faster and more accurate recognition software, including real-time face detection in video footage and emotional recognition.
Today, facial recognition technology is used in a variety of applications. It is often a prominent feature in social media platforms, such as Facebook, Snapchat and TikTok. For instance, DeepFace, a “deep learning” facial recognition system created by Facebook, helps the platform identify photos of users so they can review or share the content. (See Heilweil, Facebook is backing away from facial recognition. Meta isn’t. Vox.com (Nov. 3, 2021) https://www.vox.com/recode/22761598/facebook-facial-recognition-meta> [last visited Apr. 5, 2023].) Snapchat employs similar technology to allow users to share content augmented by “filters,” which can add features or alter an image of the user’s face. Facial recognition technology has also seen increasing use as a method of ID verification, such as with Apple’s Face ID and Google’s Android “Ice Cream Sandwich” systems.
As facial recognition technology has become more widespread, so have concerns about its shortcomings and potential for misuse. Many critics highlight that the use of facial recognition systems result in serious privacy violations, and that mechanisms to protect against the unwanted sale or dissemination of personal biometric data are insufficient. (Schwartz, Resisting the Menace of Face Recognition, Electronic Frontier Foundation (Oct. 26, 2021) <https://www.eff.org/deeplinks/2021/10/resisting-menace-face-recognition> [last visited Apr. 5, 2023] Others suggest that the technology is still too inaccurate and unreliable to be used in such a broad array of applications. For instance, facial recognition technology may not be less accurate when identifying transgender people, women, or persons with darker complexions. (See e.g., Facial Recognition Software Has a Gender Problem, National Science Foundation (Nov. 1, 2019) <https://www.nsf.gov/discoveries/disc_summ.jsp?cntn_id=299486>[last visited Apr. 5, 2023]; Buolamwini et al., Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification PMLR 81:77-91, 2018 <http://proceedings.mlr.press/v81/buolamwini18a/buolamwini18a.pdf>[last visited Apr. 5, 2023] (Najibi, Racial Discrimination in Face Recognition Technology, Harvard University Graduate School of Arts and Sciences Blog (Oct. 24, 2020) <https://sitn.hms.harvard.edu/flash/2020/racial-discrimination-in-face-recognition-technology/> [last visited Apr. 5, 2023].) - Law Enforcement Uses of Facial Recognition Systems: Despite growing concerns, law enforcement agencies at the federal, state and local level continue to use facial recognition programs. A Government Accountability Office report revealed that 20 federal agencies employ such programs, 10 of which intend to expand them over the coming years. (Facial Recognition Technology: Federal Law Enforcement Agencies Should Better Assess Privacy and Other Risks, United States Government Accountability Office. (June 3, 2021) <https://www.gao.gov/products/gao-21-518> [last visited Apr. 5, 2023].) One study found that one in four law enforcement agencies across the country can access some form of FRT, and that half of American adults – more than 117 million people – are in a law enforcement face recognition network. (Garvie et al., The Perpetual Line-Up: Unregulated Police Face Recognition in America, Georgetown Law Center on Privacy and Technology (Oct. 18, 2016) <https://www.perpetuallineup.org/> [last visited Apr. 5, 2023].) Very few of these agencies have a formal facial recognition policy, but one such agency, the New York Police Department, defines the scope of its policy as follows: “Facial recognition technology enhances the ability to investigate criminal activity and increases public safety. The facial recognition process does not by itself establish probable cause to arrest or obtain a search warrant, but it may generate investigative leads through a combination of automated biometric comparisons and human analysis.” (Facial Recognition Technology Patrol Guide, City of New York Police Department (Mar. 12, 2020) <https://www1.nyc.gov/assets/nypd/downloads/pdf/nypd-facial-recognition-patrol-guide.pdf> [last visited Apr. 5, 2023].) Proponents of facial recognition technology see it as a useful tool in helping identify criminals. It was reportedly utilized to identify the man charged in the deadly shooting at The Capital Gazette’s newsroom in Annapolis, Maryland in 2018. (Singer, Amazon’s Facial Recognition Wrongly Identifies 28 Lawmakers, A.C.L.U. Says, New York Times, (July 26, 2018) <https://www.nytimes.com/2018/07/26/technology/amazon-aclu-facial-recognition-congress.html?login=facebook> [last visited Apr. 5, 2023].)
The inaccuracy, biases, and potential privacy intrusions inherent in many facial recognition systems used by law enforcement have led to criticism from civil rights advocates, especially in California. In March 2020, the ACLU, on behalf of a group of California residents, filed a class action lawsuit against Clearview AI, claiming that the company illegally collected biometric data from social media and other websites, and applied facial recognition software to the databases for sale to law enforcement and other companies. (Clearview AI class-action may further test CCPA’s private right of action, JD Supra (Mar. 12, 2020) <https://www.jdsupra.com/legalnews/clearview-ai-class-action-may-further-14597/> [last visited Apr. 5, 2023].) An investigation by Buzzfeed in 2021 found that 140 state and local law enforcement agencies in California had used or tried Clearview AI’s system. (Mac et al., Your Local Police Department Might Have Used This Facial Recognition Tool To Surveil You. Find Out Here. Buzzfeed News, (Apr. 6, 2021) <https://www.buzzfeednews.com/article/ryanmac/facial-recognition-local-police-clearview-ai-table> [last visited Apr. 5, 2023].)
The controversy surrounding law enforcement use of facial recognition has led many California cities to ban the technology, including San Francisco, Oakland, Berkeley, Santa Cruz and Alameda. Despite the ban in San Francisco, officers there may have skirted the city’s ban by outsourcing an FRT search to another law enforcement agency. (Cassidy, Facial recognition tech used to build SFPD gun case, despite city ban, San Francisco Chronicle (Sept. 24, 2020) < https://www.sfchronicle.com/bayarea/article/Facial-recognition-tech-used-to-build-SFPD-gun-15595796.php> [last visited Mar. 23, 2023].)
In September 2021, the Los Angeles Times reported that the Los Angeles Police Department had used facial recognition software nearly 30,000 times since 2009, despite years of “vague and contradictory information” from the department “about how and whether it uses the technology.” According to the Times, “The LAPD has consistently denied having records related to facial recognition, and at times denied using the technology at all.” Responding to the report, the LAPD claimed that the denials were just mistakes, and that it was no secret that the department used such technology. Although the department could not determine how many leads from the system developed into arrests, it asserted that “the technology helped identify suspects in gang crimes where witnesses were too fearful to come forward and in crimes where no witnesses existed.” (Rector et al., Despite past denials, LAPD has used facial recognition software 30,000 times in last decade, records show, Los Angeles Times, (Sept. 21, 2020) https://www.latimes.com/california/story/2020-09-21/lapd-controversial-facial-recognition-software.) - Facial Recognition Technology Legislation in California: In 2019, the Legislature passed Assembly Bill 1215 (Ting), Chapter 579, Statutes of 2019, which banned the use of facial recognition technology and other biometric surveillance systems in connection with cameras worn or carried by law enforcement, including body-worn cameras (BWC), for the purpose of identifying individuals using biometric data. This ban covered both the direct use of biometric surveillance by a law enforcement officer or agency, as well as a request or agreement by an officer or agency that another officer or agency, or a third party, use a biometric surveillance system on behalf of the requesting party. The ban also included narrow exceptions for processes that redact a recording prior to disclosure in order to protect the privacy of a subject, and the use of a mobile fingerprint-scanning device to identify someone without proof of identification during a lawful detention, as long as neither of these functions result in the retention of biometric data or surveillance information. AB 1215 included a sunset date of January 1, 2023.
SB 1038 (Bradford), of the 2021-2022 Legislature, would have extended the ban on biometric surveillance and facial recognition systems in connection with cameras worn or carried by officers indefinitely. At its core, the question involved balancing the purported investigatory benefits of facial recognition technology against its demonstrated privacy risks, technical flaws and racial and gender biases. Committee staff did not identify or receive any evidence demonstrating that the ban on facial recognition technology used in connection with BWC had significantly hampered law enforcement efforts in the two years since it became operative. (Sen. Com. on Public Saf., com. on Sen. Bill No. 1038 (2021-2022 Reg. Sess.).) SB 1038 failed passage in the Senate.
This year, the Legislature is asked once again to determine whether the investigatory benefits of facial recognition technology outweigh the risk to the communities served by law enforcement. This bill would set minimum standards for use of FRT by law enforcement, including requiring law enforcement agencies to have a written policy for FRT use, allowing for FRT use when a peace officer has reasonable suspicion that an individual has committed a felony, and providing that an FRT-generated match of an individual may not be the sole basis for probable cause for an arrest, search, or affidavit for a warrant. It does not include any limitation on the source of the input image submitted for comparison against the database of persons. Police could use traffic cameras, CCTV, and images from BWCs or dashcams.
In contrast, AB 1034 (Wilson) would prohibit a law enforcement officer or agency from installing, activating, or using a biometric surveillance system solely in connection with a law enforcement agency’s body-worn camera or any other camera. This bill would allow for input images from more sources than AB 1034 would ban—the two bills are reconcilable. - Reasonable Suspicion to Use FRT: This bill provides that law enforcement may use FRT when they have reasonable suspicion to believe that the person either has committed or is committing a crime. Reasonable suspicion exists “when the detaining officer can point to specific articulable facts that, considered in light of the totality of the circumstances, provide some objective manifestation that the person detained may be involved in criminal activity.” (People v. Souza (1994) 9 Cal.4th 224, 231, citing U.S. v. Cortez (1981) 449 U.S. 411, 417-418.) If, during the detention, a reasonable person would believe that the individual is armed and dangerous, the officer may also conduct a brief patdown (or “frisk”) of an individual’s outer clothing for weapons. (Terry v. Ohio (1968) 392 U.S. 1, 21.). A brief detention based on reasonable suspicion, and the possible subsequent patdown for weapons, is known as a Terry stop.
The U.S. Supreme Court has not yet decided whether law enforcement may require an individual produce identification during a Terry stop. Law enforcement may request that a person stopped based on reasonable suspicion disclose their name during the stop. (See Hiibel v. Sixth Judicial Dist. Court (2004) 542 U.S. 177, 185.) However, a suspect is not required to provide identification to law enforcement during a Terry stop as a matter of law. In California, the authority to require a suspect to produce identification is limited. (See Lawson v. Kolender (9th Cir. 1981) 658 F.2d 1362; Martinelli v. Beaumont (9th Cir. 1987) 820 F.2d 1491, 1494; People v. Garcia (2006) 145 Cal.App.4th 782, 788.) This bill would circumvent those limitations by allowing for FRT whenever a peace officer has reasonable suspicion that a person has committed a felony.
Moreover, there is little question about which communities are most likely to be subject to FRT use by law enforcement based on reasonable suspicion. According to a recent report, “Relative to other racial/ethnic groups, Black individuals had the highest proportion of their stops reported as reasonable suspicion.” (Racial and Identity Profiling Advisory Board (RIPA), Annual Report 2023 (Jan. 1, 2023) p. 8 <https://oag.ca.gov/system/files/media/ripa-board-report-2023.pdf> [last visited Mar. 23, 2023].) Both transgender men/boys and transgender women/girls reported substantially higher proportions of stops as reasonable suspicion—ranging from 42.9% to 45.4% of the total reported stops, respectively—than cisgender and gender nonconforming individuals (Id. at p. 40.) RIPA further found, “Individuals perceived to be LGBT had…a higher proportion of their stops reported as reasonable suspicion…than individual who officers did not perceive to be LGBT.” (Id. at p. 41.)
Given existing limitations on requiring identification based on reasonable suspicion and the disproportionate application of reasonable suspicion against marginalized communities, would requiring a higher standard (e.g. probable cause) to justify the use of FRT by law enforcement be more appropriate?
- Argument in Support: According to the League of California Cities, “Facial recognition technology is one of many tools utilized in identifying an individual by comparing a digital image of the person’s face to a database of known faces, typically by measuring distinct facial features and characteristics. This technology does not by itself result in ultimate identification, but it may generate investigative leads necessary for combatting crime within our communities. Technology assists our law enforcement partners in doing their jobs more efficiently and ultimately improves public safety.
“Cal Cities supports accountability on the part of law enforcement agencies concerning police technology and policies, as well as related oversight by local governing bodies. However, we do not support policies that restrict law enforcement agencies from utilizing technologies that would otherwise enhance their ability to prevent criminal activity in the communities they serve.” - Argument in Opposition: According to ACLU California Action, “AB 642 Places Often-Targeted Communities at Risk
“If AB 642 becomes law, California will see an explosion of the sensitive databases containing face prints and biometric information that are part and parcel of FRT. These databases are routinely and disproportionately populated with millions of mugshots of Black and Brown victims of the War on Drugs and other discriminatory laws. As a result, Black and Brown people may be more likely to be identified by systems running on this technology, profiled, stopped, detained, and entered into these databases, magnifying a cycle of racially biased policing and incarceration. Unfortunately, recent amendments to AB 642 make this problem worse, sanctioning the expansion of mugshot databases while explicitly permitting police to make use of unaccountable for-profit databases sold by data brokers and other private companies.
“The amendments to AB 642 fail to meaningfully limit the dangerous sharing of information with other agencies, risking serious consequences for women and LGBTQI people. The national attacks on abortion rights, bodily autonomy, and trans people demand that California provide a haven for those seeking reproductive or gender-affirming care. Amendments to AB 642 merely limit police from sharing info about the literal ‘the provision’ of these services, preserving wide latitude for police to share information about people who plan to seek care or have already done so and are travelling within California. AB 642 incentivizes uses of FRT that could be exploited to identify and prosecute people who travel to our state to visit a doctor’s office or a Planned Parenthood clinic.
“AB 642 Places Communities in Danger and Unjustifiably Erodes Civil Rights
“AB 642 threatens more than our privacy. It is also foreseeable that out-of-state and federal agencies will demand access to and exploit the biometric databases ushered in by AB 642, placing dangerous pressure on our state’s sanctuary laws. ICE has already been caught tapping into and demanding access to facial recognition databases in other states. Amendments to AB 642 do nothing to address the inherent vulnerability of biometric databases to outside demands. California has chosen not to proactively use state resources for immigration enforcement, and it should continue to protect immigrants.
“By allowing police to scan and identify people with few limitations, AB 642 will also increase unnecessary interactions with innocent residents. With regularity, we are learning stories about Black men like Robert Williams. Mr. Williams was wrongfully arrested in his driveway with his wife and daughter watching; he was jailed because police misused facial recognition. In Mr. Williams’ case, police ignored warnings like those contained in AB 642, so we know this bill will not prevent egregious mistakes. Indeed, recent amendments would actually prevent people like Mr. Williams from bringing a lawsuit and seeking justice wherever police claim they were acting in good faith. By legally sanctioning the widespread use of FRT, AB 642 will magnify the bias and over-policing already disproportionately harming communities of color, immigrants, and other minority groups. Too often we have seen that police labeling results in unjustified interactions that have the potential to easily escalate into fatal encounters. This will remain true regardless of how accurate FRT becomes. Thus, AB 642 also puts lives at risk.
“AB 642 Ignores the Evidence
“FRT is an inherently dangerous form of surveillance that gives governments the power to automatically identify us without our consent and track where we go, who we know, and even how we feel, and there is no way to prevent this encroachment on personal privacy. Rather than meaningfully constrain this power, amendments to AB 642 explicitly specify that police may use FRT for ‘real time’ surveillance of entire areas, exposing entire communities to the scanning of their faces and identities by police cameras mounted on streetlights, buildings, and attached to officers. Simply put, this is how authoritarian states use this technology to track, target, and control their populations and marginalized communities.
“As a result, the face surveillance systems permitted by AB 642 will chill the exercise of important civil rights, including at protests, places of worship, and political gatherings. There are no acceptable standards under which law enforcement can use face surveillance. Rather than aligning with the civil rights community and national consensus, AB 642 grants law enforcement agencies sweeping statutory authority to use face recognition technology to identify and track people across the state.
“There is a strong and growing public consensus that FRT is simply too dangerous and corrosive to our rights to be used by law enforcement. Companies like Amazon, Microsoft, and IBM refuse to sell FRT to police, as has Axon, the most prominent police body camera maker. At least 20 U.S. cities, including your hometown of San Francisco, have banned the government use of face recognition technology. Progressive leadership in the United States Congress recently introduced a bill that would prohibit the government’s use of facial recognition and condition funding to localities on their adopting the same. These local laws and federal bill recognize that the most responsible standard for FRT is a ban on its use by governments. Prohibitions on the government use of facial recognition protect our privacy, reduce dangerous encounters and wrongful detentions, safeguard our freedom of speech, and impede the creation of dangerous biometric databases.” - Related Legislation:
- AB 1034 (Wilson), would prohibit a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other officer camera. AB 1034 is pending hearing in the Assembly Committee on Privacy and Consumer Protection.
- AB 79 (Weber), would prohibit a peace officer from using deadly force against or intending to injure, intimidate, or disorient a person by utilizing any unmanned, remotely piloted, powered ground or flying equipment except under specified circumstances. AB 79 is pending hearing in this committee.
- AB 793 (Bonta), would provide that a government entity may not seek, from any court, a compulsory process to enforce a reverse-location demand or a reverse-keyword demand, as defined. AB 793 is pending hearing in this committee.
- AB 742 (Jackson), would prohibit the use of canines by peace officers for arrest and apprehension, or in any circumstances to bite a person, but permits their use of canines for search and rescue, explosives detection, and narcotics detection. AB 742 is pending hearing in the Assembly Appropriations Committee.
- AB 1034 (Wilson), would prohibit a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other officer camera. AB 1034 is pending hearing in the Assembly Committee on Privacy and Consumer Protection.
- Prior Legislation:
- SB 1038 (Bradford), would have deleted the January 1, 2023 sunset date on provisions of law that prohibit a law enforcement officer from installing, activating or using a biometric surveillance system in connection with a body-worn camera or data collected by a body-worn camera. SB 1038 died on the inactive file in the Senate.
- AB 1281 (Chau), Chapter 268, Statutes of 2020, would require a business in California that uses facial recognition technology to disclose that usage in a physical sign that is clear and conspicuous at the entrance of every location that uses facial recognition technology.
- AB 1215 (Ting), Chapter 579, Statutes of 2019, prohibited a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other camera.
- SB 21 (Hill), of the 2017-2018 Legislative Session, would have required local law enforcement agencies to have a policy, approved by the local governing body, in place before using surveillance technology. SB 21 was held in the Assembly Appropriations Committee.
- AB 69 (Rodriguez) Chapter 461, Statutes of 2015, requires law enforcement agencies to follow specified best practices when establishing policies and procedures for downloading and storing data from body-worn cameras.
- SB 34 (Hill) Chapter 532, Statutes of 2015, imposed a variety of security, privacy and public hearing requirements on the use of automated license plate recognition systems, as well as a private right of action and provisions for remedies.
- SB 1038 (Bradford), would have deleted the January 1, 2023 sunset date on provisions of law that prohibit a law enforcement officer from installing, activating or using a biometric surveillance system in connection with a body-worn camera or data collected by a body-worn camera. SB 1038 died on the inactive file in the Senate.
REGISTERED SUPPORT / OPPOSITION:
Support
League of California Cities
Orange County Board of Supervisors - Supervisor Vicente Sarmiento
Security Industry Association
Opposition
A New Way of Life
Access Reproductive Justice
ACLU California Action
All of Us or None Riverside
Anti Police-terror Project
Asian Americans Advancing Justice - Asian Law Caucus
Asian Law Alliance
Bend the Arc California
California Association of Black Lawyers
California Coalition for Women Prisoners
California Innocence Coalition: Northern California Innocence Project, California Innocence Project, Loyola Project for The Innocent
California Public Defenders Association (CPDA)
California United for A Responsible Budget (CURB)
Care First California
Central American Resource Center of San Francisco
Citizens for A Better Los Angeles
Coalition for Homelessness San Francisco
Electronic Frontier Foundation
Equal Justice Society
Family Reunification Equity & Empowerment (F.R.E.E.)
Free Speech Coalition
If/when/how: Lawyering for Reproductive Justice
Indivisible CA Statestrong
Indivisible Ca: Statestrong
Initiate Justice
Initiate Justice Action
Lawyers Committee for Civil Rights of The San Francisco Bay Area
Lawyers' Committee for Civil Rights of The San Francisco Bay Area
Legal Services for Prisoners With Children
Long Beach Immigrant Rights Coalition
Media Alliance
Mediajustice
National Center for Lesbian Rights
Oakland Privacy
Orange County Rapid Response Network
People's Budget Orange County
Positive Women's Network - USA
Privacy Rights Clearinghouse
Safer Streets LA
San Francisco Public Defender - Racial Justice Committee
Secure Justice
St James Infirmary
St. James Infirmary
Starting Over INC.
Stop the Musick Coalition
Tenth Amendment Center
Training in Early Abortion for Comprehensive Health Care
Transforming Justice Orange County
Transgender, Gendervariant, Intersex Justice Project
Urge: Unite for Reproductive & Gender Equity
Analysis Prepared by:
Date of Hearing: March 28, 2023
Counsel: Andrew Ironside
ASSEMBLY COMMITTEE ON
Reginald Byron Jones-Sawyer, Sr., Chair
ABPCA Bill Id:
Author:() – As Ver:
SUMMARY:
Sets minimum standards for use of facial recognition technology (FRT) by law enforcement, including requiring law enforcement agencies to have a written policy for FRT use, allowing for FRT use to identify both individuals who are suspects in a crime and those that are not, and providing that an FRT-generated match of an individual may not be the sole basis for probable cause for an arrest, search, or affidavit for a warrant. Specifically, this bill:
- Requires a law enforcement agency that operates FRT to meet both of the following requirements commencing on July 1, 2024:
- Only FRT systems with algorithms that have been evaluated under the National Institute of Standards and Technology Face Recognition Vender Test Program and have demonstrated an accuracy score of at least 98% true positives within two or more data sets relevant to investigative applications on a program report shall be used; and
- Requires the agency to have a written policy for FRT use, as specified.
- Only FRT systems with algorithms that have been evaluated under the National Institute of Standards and Technology Face Recognition Vender Test Program and have demonstrated an accuracy score of at least 98% true positives within two or more data sets relevant to investigative applications on a program report shall be used; and
- Requires the agency to have a written policy that includes, without limitation, all of the following:
- A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT from the Commission on Peace Officer Standards and Training;
- A requirement that a manager be assigned to oversee the FRT program;
- A policy that describes the parameters of acceptable inputs to be used for queries of available databases and that prohibits the use of sketches or other manually produced images;
- An acceptable use policy that includes specific allowances and restriction on use for all of the following:
- To identify a suspect in alleged criminal behavior where reasonable suspicion exists that a crime has been or is being committed and the person whose image is being analyzed is the person who has committed or is committing that crime;
- To identify a victim or witness to a crime;
- To identify an unidentified deceased person;
- To identify a person who is missing, incapacitated, or unable to identify themselves;
- To identify a person who is lawfully detained and has not produced valid identification;
- To investigate a credible threat of violence; and,
- To mitigate an imminent threat to public safety.
- To identify a suspect in alleged criminal behavior where reasonable suspicion exists that a crime has been or is being committed and the person whose image is being analyzed is the person who has committed or is committing that crime;
- A prohibited use policy that prohibits both of the following:
- The use of FRT to identify an individual solely because of their race, color, religious beliefs, sexual orientation, gender, disability, national origin, or membership in any other class protected by law against discrimination; and
- The use of FRT to identify an individual solely engaged in the exercise of a constitutionally protected right, including, without limitation, speech, public assembly, or the practice of religion, when the person has not violated any law, unless necessary to identify a victim or witness of a serious or violent felony, or to defend against an imminent or immediate threat to death or serious bodily injury.
- The use of FRT to identify an individual solely because of their race, color, religious beliefs, sexual orientation, gender, disability, national origin, or membership in any other class protected by law against discrimination; and
- A requirement that a record of all FRT queries be maintained by the agency.
- A requirement that FRT use be limited to specifically authorized personnel who have received certified training in the use of FRT from the Commission on Peace Officer Standards and Training;
- Provides that FRT is an investigative tool to be used as an aid in identifying persons and generating investigative leads.
- Provides that an FRT-generated match shall not, under any circumstances, provide the sole basis for probable cause for an arrest, search, or affidavit for a warrant.
- Requires any law enforcement agency that uses FRT to post both of the following on their internet website:
- A copy of the agency’s FRT policy, as specified; and,
- Commencing on January 1, 2025, and annually thereafter, an annual report summarizing FRT usage for the previous calendar year.
- A copy of the agency’s FRT policy, as specified; and,
- Provides that the admissibility of an FRT query result as evidence in any court proceeding is governed by existing law.
- Defines “facial recognition technology” or “FRT” as a system that compares an input image of an unidentified human face against a database of known persons and, based on biometric data, generates possible matches to aid in identifying the person in the input image.
- Provides that FRT does not include any access control system used by a law enforcement agency that uses biometric inputs to confirm the identity of employees or other approved persons for the purpose of controlling access to any secured place, device, or system.
- Defines “law enforcement agency” as any department or agency of the state or any political subdivision thereof that employs any peace officer, as defined.
EXISTING LAW:
- Declares that it is the intent of the Legislature to establish policies and procedures to address issues related to the downloading and storage data recorded by a body-worn camera worn by a peace officer; these policies and procedures shall be based on best practices. (Pen. Code, § 832.18, subd. (a).)
- Encourages agencies to consider best practices in establishing when data should be downloaded to ensure the data is entered into the system in a timely manner, the cameras are properly maintained and ready for the next use, and for purposes of tagging and categorizing the data. (Pen. Code, § 832.18, subd. (b).)
- Encourages agencies to consider best practices in establishing specific measures to prevent data tampering, deleting, and copying, including prohibiting the unauthorized use, duplication, or distribution of body-worn camera data. (Pen. Code, § 832.18, subd. (b)(3).)
- Encourages agencies to consider best practices in establishing the length of time that recorded data is to be stored. States that nonevidentiary data including video and audio recorded by a body-worn camera should be retained for a minimum of 60 days, after which it may be erased, destroyed, or recycled. Provides that an agency may keep data for more than 60 days to have it available in case of a civilian complaint and to preserve transparency. (Pen. Code, § 832.18, subd. (b)(5)(A).)
- States that evidentiary data including video and audio recorded by a body-worn camera should be retained for a minimum of two years under any of the following circumstances:
- The recording is of an incident involving the use of force by a peace officer or an officer-involved shooting;
- The recording is of an incident that leads to the detention or arrest of an individual; or,
- The recording is relevant to a formal or informal complaint against a law enforcement officer or a law enforcement agency. (Pen. Code, § 832.18, subd. (b)(5)(B).)
- States that the recording should be retained for additional time as required by law for other evidence that may be relevant to a criminal prosecution. (Pen. Code, § 832.18, subd. (b)(5)(C).)
- Instructs law enforcement agencies to work with legal counsel to determine a retention schedule to ensure that storage policies and practices are in compliance with all relevant laws and adequately preserve evidentiary chains of custody. (Pen. Code, § 832.18, subd. (b)(5)(D).)
- Encourages agencies to adopt a policy that records or logs of access and deletion of data from body-worn cameras should be retained permanently. (Pen. Code, § 832.18, subd. (b)(5)(E).)
- Encourages agencies to include in a policy information about where the body-worn camera data will be stored, including, for example, an in-house server which is managed internally, or an online cloud database which is managed by a third-party vendor. (Pen. Code, § 832.18, subd. (b)(6).)
- Instructs a law enforcement agency using a third-party vendor to manage the data storage system, to consider the following factors to protect the security and integrity of the data: Using an experienced and reputable third-party vendor; entering into contracts that govern the vendor relationship and protect the agency’s data; using a system that has a built-in audit trail to prevent data tampering and unauthorized access; using a system that has a reliable method for automatically backing up data for storage; consulting with internal legal counsel to ensure the method of data storage meets legal requirements for chain-of-custody concerns; and using a system that includes technical assistance capabilities. (Pen. Code, § 832.18, subd. (b)(7).)
- Encourages agencies to include in a policy a requirement that all recorded data from body-worn cameras are property of their respective law enforcement agency and shall not be accessed or released for any unauthorized purpose. Encourages a policy that explicitly prohibits agency personnel from accessing recorded data for personal use and from uploading recorded data onto public and social media Internet websites, and include sanctions for violations of this prohibition. (Pen. Code, § 832.18, subd. (b)(8).)
- Requires that a public agency that operates or intends to operate an Automatic License Plate Recognition (ALPR) system to provide an opportunity for public comment at a public meeting of the agency's governing body before implementing the program. (Civil Code, § 1798.90.55.)
- Prohibits a local agency from acquiring cellular communications interception technology unless approved by its legislative body. (Gov. Code, § 53166, subd. (c)(1).)
FISCAL EFFECT:
Unknown
COMMENTS:
- Author's Statement: According to the author, “I authored AB 1215 in 2019 which banned the use of biometric surveillance through police body cameras. The bill only passed with a three year moratorium that expired January 1, 2023. Consequently, current law has absolutely no parameters set regarding law enforcement’s use of facial recognition technology. It is critical that we ensure there are safeguards in place in order to avoid another year of unregulated use. We can’t go another year with no protections. AB 642 is a response to a battle that we cannot afford to risk losing. The bill includes critical safeguards such as codifying an accuracy level, oversight and reporting, and prohibits law enforcement from using a match alone to arrest someone, to request a warrant, to violate someone’s constitutional rights, and to discriminate against protected characteristics. Most importantly, this bill does not prohibit nor deter local governments from choosing to ban the use of facial recognition technology.”
- Facial Recognition Technology: Facial recognition technology is capable of identifying an individual by comparing a digital image of the person’s face to a database of known faces, typically by measuring distinct facial features and characteristics. Early versions of the technology were pioneered in the 1960s and 1970s, but true facial recognition technology as we understand it today did not come about until the early 1990s. In 1993, the United States military developed the Facial Recognition Technology (FERET) program, which aimed to create a database of faces and recognition algorithms to assist in intelligence gathering, security and law enforcement. (“Facial Recognition Technology (FERET).”) The National Institute of Standards and Technology, United States Department of Commerce. https://www.nist.gov/programs-projects/face-recognition-technology-feret).) Since that time, advances in computer technology and machine learning have led to faster and more accurate recognition software, including real-time face detection in video footage and emotional recognition.
Today, facial recognition technology is used in a variety of applications. It is often a prominent feature in social media platforms, such as Facebook, Snapchat and TikTok. For instance, DeepFace, a “deep learning” facial recognition system created by Facebook, helps the platform identify photos of users so they can review or share the content. (See “Facebook is backing away from facial recognition. Meta isn’t.” (Nov. 3, 2021) https://www.vox.com/recode/22761598/facebook-facial-recognition-meta.) Snapchat employs similar technology to allow users to share content augmented by “filters,” which can add features or alter an image of the user’s face. Facial recognition technology has also seen increasing use as a method of ID verification, such as with Apple’s Face ID and Google’s Android “Ice Cream Sandwich” systems.
As facial recognition technology has become more widespread, so have concerns about its shortcomings and potential for misuse. Many critics highlight that the use of facial recognition systems result in serious privacy violations, and that mechanisms to protect against the unwanted sale or dissemination of personal biometric data are insufficient. (Schwartz, Resisting the Menace of Face Recognition, Electronic Frontier Foundation (Oct. 26, 2021) https://www.eff.org/deeplinks/2021/10/resisting-menace-face-recognition.) Others suggest that the technology is still too inaccurate and unreliable to be used in such a broad array of applications. For instance, studies suggest that while facial recognition systems have had increasing success identifying cis-gendered individuals, these systems get it wrong more than one-third of the time if the face belongs to a transgender person. (Facial Recognition Software Has a Gender Problem, National Science Foundation (Nov. 1, 2019) https://www.nsf.gov/discoveries/disc_summ.jsp?cntn_id=299486.) However, even among cis-gendered individuals, research shows that facial recognition systems can be significantly less accurate when identifying women than when identifying men. (Buolamwini et al., Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification PMLR 81:77-91, 2018 http://proceedings.mlr.press/v81/buolamwini18a/buolamwini18a.pdf.) Additionally, a growing body of research demonstrates that facial recognition systems are significantly less accurate in identifying individuals with dark complexions, particularly women (Najibi, Racial Discrimination in Face Recognition Technology, Harvard University Graduate School of Arts and Sciences Blog (Oct. 24, 2020) https://sitn.hms.harvard.edu/flash/2020/racial-discrimination-in-face-recognition-technology/.) - Law Enforcement Uses of Facial Recognition Systems: Despite growing concerns, law enforcement agencies at the federal, state and local level continue to use facial recognition programs. A recent Government Accountability Office report revealed that 20 federal agencies employ such programs, 10 of which intend to expand them over the coming years. (Facial Recognition Technology: Federal Law Enforcement Agencies Should Better Assess Privacy and Other Risks, United States Government Accountability Office. (June 3, 2021) https://www.gao.gov/products/gao-21-518.) One study found that one in four law enforcement agencies across the country can access some form of FRT, and that half of American adults – more than 117 million people – are in a law enforcement face recognition network. (Garvie et al., The Perpetual Line-Up: Unregulated Police Face Recognition in America, Georgetown Law Center on Privacy and Technology (Oct. 18, 2016) https://www.perpetuallineup.org/.) Very few of these agencies have a formal facial recognition policy, but one such agency, the New York Police Department, defines the scope of its policy as follows: “Facial recognition technology enhances the ability to investigate criminal activity and increases public safety. The facial recognition process does not by itself establish probable cause to arrest or obtain a search warrant, but it may generate investigative leads through a combination of automated biometric comparisons and human analysis.” (Facial Recognition Technology Patrol Guide, City of New York Police Department (Mar. 12, 2020) https://www1.nyc.gov/assets/nypd/downloads/pdf/nypd-facial-recognition-patrol-guide.pdf.) Proponents of facial recognition technology see it as a useful tool in helping identify criminals. It was reportedly utilized to identify the man charged in the deadly shooting at The Capital Gazette’s newsroom in Annapolis, Maryland in 2018. (Singer, Amazon’s Facial Recognition Wrongly Identifies 28 Lawmakers, A.C.L.U. Says, New York Times, (July 26, 2018) https://www.nytimes.com/2018/07/26/technology/amazon-aclu-facial-recognition-congress.html?login=facebook.)
The inaccuracy, biases and potential privacy intrusions inherent in many facial recognition systems used by law enforcement have led to criticism from civil rights advocates, especially in California. In March 2020, the ACLU, on behalf of a group of California residents, filed a class action lawsuit against Clearview AI, claiming that the company illegally collected biometric data from social media and other websites, and applied facial recognition software to the databases for sale to law enforcement and other companies. (Clearview AI class-action may further test CCPA’s private right of action, JD Supra (Mar. 12, 2020) https://www.jdsupra.com/legalnews/clearview-ai-class-action-may-further-14597/.) An investigation by Buzzfeed in 2021 found that 140 state and local law enforcement agencies in California had used or tried Clearview AI’s system. (Your Local Police Department Might Have Used This Facial Recognition Tool To Surveil You. Find Out Here. Buzzfeed News, (Apr. 6, 2021) https://www.buzzfeednews.com/article/ryanmac/facial-recognition-local-police-clearview-ai-table.)
The controversy surrounding law enforcement use of facial recognition has led many California cities to ban the technology, including San Francisco, Oakland, Berkeley, Santa Cruz and Alameda. Despite the ban in San Francisco, officers there may have skirted the city’s ban by outsourcing an FRT search to another law enforcement agency. (Cassidy, Facial recognition tech used to build SFPD gun case, despite city ban, San Francisco Chronicle (Sept. 24, 2020) < https://www.sfchronicle.com/bayarea/article/Facial-recognition-tech-used-to-build-SFPD-gun-15595796.php> [last visited Mar. 23, 2023].)
In September 2021, the Los Angeles Times reported that the Los Angeles Police Department had used facial recognition software nearly 30,000 times since 2009, despite years of “vague and contradictory information” from the department “about how and whether it uses the technology.” According to the Times, “The LAPD has consistently denied having records related to facial recognition, and at times denied using the technology at all.” Responding to the report, the LAPD claimed that the denials were just mistakes, and that it was no secret that the department used such technology. Although the department could not determine how many leads from the system developed into arrests, it asserted that “the technology helped identify suspects in gang crimes where witnesses were too fearful to come forward and in crimes where no witnesses existed.” (Despite past denials, LAPD has used facial recognition software 30,000 times in last decade, records show, Los Angeles Times, (Sept. 21, 2020) https://www.latimes.com/california/story/2020-09-21/lapd-controversial-facial-recognition-software.) - Facial Recognition Technology Legislation in California: In 2019, the Legislature passed Assembly Bill 1215 (Ting), Chapter 579, Statutes of 2019, which banned the use of facial recognition technology and other biometric surveillance systems in connection with cameras worn or carried by law enforcement, including body-worn cameras (BWC), for the purpose of identifying individuals using biometric data. This ban covered both the direct use of biometric surveillance by a law enforcement officer or agency, as well as a request or agreement by an officer or agency that another officer or agency, or a third party, use a biometric surveillance system on behalf of the requesting party. The ban also included narrow exceptions for processes that redact a recording prior to disclosure in order to protect the privacy of a subject, and the use of a mobile fingerprint-scanning device to identify someone without proof of identification during a lawful detention, as long as neither of these functions result in the retention of biometric data or surveillance information. AB 1215 included a sunset date of January 1, 2023.
SB 1038 (Bradford), of the 2021-2022 Legislature, would have extended the ban on biometric surveillance and facial recognition systems in connection with cameras worn or carried by officers indefinitely. At its core, the question involved balancing the purported investigatory benefits of facial recognition technology against its demonstrated privacy risks, technical flaws and racial and gender biases. Committee staff did not identify or receive any evidence demonstrating that the ban on facial recognition technology used in connection with BWC had significantly hampered law enforcement efforts in the two years since it became operative. (Sen. Com. on Public Saf., com. on Sen. Bill No. 1038 (2021-2022 Reg. Sess.).) SB 1038 failed passage in the Senate.
This year, the Legislature is asked once again to determine whether the investigatory benefits of facial recognition technology outweigh the risk to the communities served by law enforcement. This bill would set minimum standards for use of facial recognition technology by law enforcement, including requiring a law enforcement agency to have written policy governing FRT use, allowing for FRT use to identify both individuals who are suspects in a crime and those that are not, and providing that an FRT-generated match of an individual may not be the sole basis for probable cause for an arrest, search, or affidavit for a warrant. It does not include any limitation on the source of the input image submitted for comparison against the database of persons. Police could use traffic cameras, CCTV, and images from BWCs or dashcams.
In contrast, AB 1034 (Wilson) would prohibit a law enforcement officer or agency from installing, activating, or using a biometric surveillance system solely in connection with a law enforcement agency’s body-worn camera or any other camera. This bill would allow for input images from more sources than AB 1034 would ban—the two bills are reconcilable. - Broad Application: This bill would require that law enforcement agencies develop acceptable use policies that include specific allowances and restrictions on use for certain people or under broad circumstances. However, the use policy provision states that those specific allowances and restrictions on use are not limited to the persons and circumstances identified in the bill—the list is “without limitation.” This bill would prohibit use in only two circumstances: to identify an individual solely because their race color, religious beliefs, sexual orientation, gender, disability, national origin, or membership in any other protected class; or to identify an individual engaged in the exercise of a constitutionally protected right.
Other than those prohibitions, there is nothing to prevent a law enforcement agency from expanding use to include people not identified in the bill. The bill already allows FRT use to identify people not even suspected of any criminal conduct, such as victims and witnesses. What is to prevent a law enforcement agency from developing, in the name of generating investigative leads, an acceptable use policy for FRT to identify the friends and family of a person who police have reasonable suspicion to believe has committed a crime?
Moreover, this bill provides that law enforcement may use FRT to identify a suspect if an officer has reasonable suspicion that a crime has been or is being committed. It also allows law enforcement to use FRT to identify the victim or witness to a crime. Under current law, “crime” means “an act committed or omitted in violation of a law forbidding or commanding it,” which results in death, imprisonment, fine, or removal from office. (Pen. Code, § 15.) A crime could be a felony, a misdemeanor, or an infraction. It could be a violation of state law or local code or ordinance. Therefore, under this bill, law enforcement could use FRT to identify a person that an officer has reasonable suspicion to believe has taken more than 25 copies of the current issue of a free or complimentary newspaper. (Pen. Code, §§ 187, 647, 490.7, subd. (b).)
Law enforcement could use FRT to identify the witnesses to and victims of those crimes, too. Under this bill, could law enforcement use facial recognition to identify the publisher of a free or complimentary newspaper who has had more than 25 copies of a current issue taken by someone? If so, would that implicate the First Amendment right to freedom of speech and freedom of the press? - Reasonable Suspicion to Use FRT: This bill provides that law enforcement may use FRT when they have reasonable suspicion to believe that the person either has committed or is committing a crime. Reasonable suspicion exists “when the detaining officer can point to specific articulable facts that, considered in light of the totality of the circumstances, provide some objective manifestation that the person detained may be involved in criminal activity.” (People v. Souza (1994) 9 Cal.4th 224, 231, citing U.S. v. Cortez (1981) 449 U.S. 411, 417-418.) If, during the detention, a reasonable person would believe that the individual is armed and dangerous, the officer may also conduct a brief patdown (or “frisk”) of an individual’s outer clothing for weapons. (Terry v. Ohio (1968) 392 U.S. 1, 21.). A brief detention based on reasonable suspicion, and the possible subsequent patdown for weapons, is known as a Terry stop.
The U.S. Supreme Court has not yet decided whether law enforcement may require an individual produce identification during a Terry stop. Law enforcement may request that a person stopped based on reasonable suspicion disclose their name during the stop. (See Hiibel v. Sixth Judicial Dist. Court (2004) 542 U.S. 177, 185.) However, a suspect is not required to provide identification to law enforcement during a Terry stop as a matter of law. In California, the authority to require a suspect to produce identification is limited.
The Ninth Circuit has ruled that requiring a person to provide identification upon request during a Terry stop violated the prohibition against unreasonable searches and seizures under the Fourth Amendment. (Lawson v. Kolender (9th Cir. 1981) 658 F.2d 1362, aff’d Kolender v. Lawson (1983) 461 U.S. 352.) Explaining the interests involved, the court said:
Although the pat down can be a degrading experience, especially when conducted in public view, it is ephemeral and, in the absence of weapons, lacks collateral consequences. The potential for subsequent police action or abuse is not materially enhanced.
In contrast, police knowledge of the identity of an individual they have deemed "suspicious" grants the police unfettered discretion to initiate or continue investigation of the person long after the detention has ended. Information concerning the stop, the arrest and the individual's identity may become part of a large scale data bank… We believe that the serious intrusion on personal security outweighs the mere possibility that identification may provide a link leading to arrest.
(Id. at p. 1368 [emphasis added].) The Ninth Circuit later held that using Penal Code section 148, which prohibits delaying or obstructing a police officer in the discharge of any duty, “to arrest a person for refusing to identify herself during a lawful Terry stop violates the Fourth Amendment’s proscription against unreasonable searches and seizures.” (Martinelli v. Beaumont (9th Cir. 1987) 820 F.2d 1491, 1494.)
Similarly, in People v. Garcia, the court ruled law enforcement may not frisk a suspect for the sole purpose of finding identification. (145 Cal.App.5th 782, 788.) The case arose after an officer stopped the defendant for riding his bike at night without a functioning headlamp, a violation of Vehicle Code section 21201(d). (Id. at p. 784.) The officer asked the defendant for identification, which the defendant denied having. (Ibid.) The officer then frisked the defendant and found methamphetamine. (Id. at pp. 785-86.) The sole justification for the frisk was the search for identification (Id. at p. 786.)
The court said that Terry “cannot be ‘morphed’ into a new rule to justify a search for ordinary evidence, here evidence of identification.” (Id. at p. 784.) It continued:
A fair reading of Terry v. Ohio…show[s] that the “frisk” allowable upon a proper showing was “only a frisk” for a dangerous weapon. It by no means authorizes a search for contraband, evidentiary material, or anything else in the absence of reasonable ground to arrest. Such a search is controlled by requirements of the Fourth Amendment, and probable cause is essential. Our own Supreme Court has unanimously so held. If stare decisis means anything (and it does) and if the word “only” means “only” (and it does), the trial court was required to grant [the motion to suppress the evidence resulting from the search]...
We will never know what could have happened [had the officer done something different]. What we do know is this: There is no legal justification for a patdown search for identification.
(Id. at p. 788 [internal citations omitted].)
These decisions demonstrate some limitations on the right of law enforcement to demand identification from a person who they have stopped based on reasonable suspicion. This bill would circumvent those limitations by allowing for FRT any time reasonable suspicion exists.
Regardless, there is little question about which communities are most likely to be subject to FRT use by law enforcement based on reasonable suspicion. According to a recent report, “Relative to other racial/ethnic groups, Black individuals had the highest proportion of their stops reported as reasonable suspicion.” (Racial and Identity Profiling Advisory Board (RIPA), Annual Report 2023 (Jan. 1, 2023) p. 8 <https://oag.ca.gov/system/files/media/ripa-board-report-2023.pdf> [last visited Mar. 23, 2023].) Both transgender men/boys and transgender women/girls reported substantially higher proportions of stops as reasonable suspicion—ranging from 42.9% to 45.4% of the total reported stops, respectively—than cisgender and gender nonconforming individuals (Id. at p. 40.) RIPA further found, “Individuals perceived to be LGBT had…a higher proportion of their stops reported as reasonable suspicion…than individual who officers did not perceive to be LGBT.” (Id. at p. 41.)
- Argument in Support: According to the League of California Cities, “Facial recognition technology is one of many tools utilized in identifying an individual by comparing a digital image of the person’s face to a database of known faces, typically by measuring distinct facial features and characteristics. This technology does not by itself result in ultimate identification, but it may generate investigative leads necessary for combatting crime within our communities. Technology assists our law enforcement partners in doing their jobs more efficiently and ultimately improves public safety.
“Cal Cities supports accountability on the part of law enforcement agencies concerning police technology and policies, as well as related oversight by local governing bodies. However, we do not support policies that restrict law enforcement agencies from utilizing technologies that would otherwise enhance their ability to prevent criminal activity in the communities they serve.” - Argument in Opposition: According to Secure Justice, “We have many concerns with the AB 642 (Ting) approach to authorizing the use of facial recognition technology (“FRT”) by law enforcement:
- The bill does not differentiate between FRT for investigation and FRT for surveillance. The bill would allow forms of generalized surveillance if police had reasonable suspicion (picture a public protest against FRT that violates some ordinance). Generalized surveillance should be banned.
- The bill substitutes reasonable suspicion for probable cause as a standard which is basically no standard at all. Along with the failure to distinguish described above, California would be greenlighting mass surveillance via FRT, were this bill to be enacted as written.
- The bill does not distinguish between use in serious crimes or any crime, which can be something like jaywalking, loitering, or blocking a street during a nonviolent civil protest.
- The bill does not define its terms well in terms of useable databases, photos, etc. Are we using arrest databases, DMV databases, commercial data brokers or web photo scrapers like Clearview AI? Something else?
- The bill does not address immigration enforcement.
- The bill does not address real time use with body cameras, drones, or Real-Time Crime Centers.
- The bill does not address Brady issues, or what information must be turned over to the defense when used in a case.
- The bill does not provide for clear remedies for misuse. There is no accountability mechanism for misuse.
- The bill does not provide for audits or quality checks. FRT evolves all the time.
- The bill will legitimize use on many innocent people who might just be witnesses to a crime, a clear invasion of our state and human right to privacy, and possibly infringing upon First Amendment protected activity.
- The bill does not require police to establish a written record justifying use. Without a warrant requirement there is no public, written notice of use. As reasonable suspicion is essentially no standard at all, law enforcement has a blank check to use FRT without having to justify the use. Although the bill does require that a record of queries be maintained, without having to justify use such a record is meaningless and will ensure no accountability. In addition, there is no requirement that the record be made available to the public for scrutiny.
- The bill fails to address basic questions, such as what happens to the data? How long do they hold it? Can they sell it? Can they keep it and connect it with other data (social media handles, etc.) to build more profiles of individuals.
- What is the process for double checking the match? Double blind? Human? We already have four false FRT matches that led to an arrest – 100% of these unlawful arrests were of Black men. What steps are taken to protect against error?
“Expanding upon some of the above red flags:
“Accuracy thresholds: The bill requires police to use FRT systems that are 98% accurate but does not require that the police actually use that accuracy threshold. Some of the early jurisdictions that adopted Amazon’s Rekognition (which had a high accuracy test result) turned down their own threshold to well below 98%. The system must be accurate, and the threshold test must be used to that accuracy level, otherwise the potential accuracy is meaningless.
“Use Limits: Although the bill provides an illustrative list of potential uses, it does not prohibit additional uses of FRT beyond those listed, aside from the use of FRT to target someone “solely” engaged in constitutionally protected activity, which is basically meaningless when there are so many state and local laws and regulations that law enforcement could point to as justifying the use of FRT when someone is legitimately exercising their First Amendment rights in a nonviolent manner.
“Reasonable suspicion: There are a couple problems with using reasonable suspicion and not probable cause (or higher). First, it means that FRT can be used without a crime being completed. If police believed kids are truant, or believed someone was intoxicated, they could use FRT (even if they were not actually truant or drunk). In most cases where FRT is currently used, police have a crime and want help identifying the person. Having probable cause of a crime must be a prerequisite for use. Second, reasonable suspicion is not a fixed standard. Under the totality of circumstances, law enforcement’s belief that criminal activity is afoot would give way too much discretion to police. Reasonable suspicion is no protection at all. Anyone running in a high crime area could have FRT used on them (running in a high crime area justifies the reasonable suspicion standard according to the Supreme Court). Probable cause at least gets you that a crime has occurred, and that police can articulate the crime and investigative need of the FRT search.
“Crime: The bill does not limit what kind of crimes FRT can be used for – jaywalking? Loitering? The number of misdemeanor and petty offenses that are categorized as “crimes” create too big a list to allow FRT use in California.
“Witnesses: The bill would allow FRT to be used on innocent people who have no guilt attached. In fact, it would be easier to do a FRT match on a non-suspect than a suspect. In practice, this means that anyone around a crime will be in the system because they might be a witness. There is no requirement that police try some other less invasive means. It will mean that everyone on the scene will be scanned by FRT.
“Lawfully detained: We could envision a scenario where FRT is used to identify someone lawfully arrested because they are going to be placed into custody and have their photo taken at booking. But lawfully detained means the individual or group or individuals were stopped based on reasonable suspicion or even consent. That would mean an officer could stop you on the street, ask for consent to investigate, and then run an FRT search. Lawfully detained could be a traffic stop or any of a series of contacts that is less than an arrest. In fact, under this language police could stop someone “consensually,” ask for identification, and then when the ID is refused, they run FRT because the individual was “lawfully detained”.
“Credible threat of violence: The use of this phrase lowers reasonable suspicion even more when there is violence alleged. For example, law enforcement has a hunch that an alleged gang member might be violent and with no reasonable suspicion or probable cause of a crime they could use FRT as the bill is written. As this phrase in (E) is separate from the “reasonable suspicion language in (A), it must have been included as an even lower standard, making a mockery of any “guardrails” being put into place with this bill.
“Imminent threat to public safety: This phrase is not defined. Almost anything could be an imminent threat to public safety without some definition. Is someone not wearing a mask as a Covid threat an imminent threat to public safety? Many would argue yes. The definition is dangerously vague.
“What the bill does not mention is equally as important as the red flags described above:
- What specifically must be reported in the annual reports?
- What are the ways to protect against racial bias and inaccuracy?
- Who audits FRT use, and will ensure compliance within the parameters of the bill and subsequent policy?
- When used as trial, how much information is provided to the defense?
- Where are the protections against generalized surveillance, protections for those seeking abortions, for immigrants, for those in their homes, or other protected places?
- What happens when it is misused (officer runs his ex-girlfriend’s new boyfriend)? What happens when it is politicized (DA runs it against supporters of a challenger)? There are no remedies.
“Furthermore, we encourage you to review a report from Georgetown Law’s Center on Privacy and Technology, which discusses in-depth the problems with using FRT in criminal investigations. Among their many findings:
“‘Relying on the vast wealth of research and knowledge already present in computer science, psychology, forensic science, and legal disciplines, its key findings are:
- As currently used in criminal investigations, face recognition is likely an unreliable source of identity evidence.
- The algorithm and human steps in a face recognition search each may compound the other’s mistakes.
- Since faces contain inherently biasing information such as demographics, expressions, and assumed behavioral traits, it may be impossible to remove the risk of bias and mistake.
- Face recognition has been used as probable cause to make arrests despite assurances to the contrary.
- Evidence derived from face recognition searches are already being used in criminal cases, and the accused have been deprived the opportunity to challenge it.
- The harms of wrongful arrests and investigations are real, even if they are hard to quantify.’”
- Related Legislation:
- AB 1034 (Wilson), would prohibit a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other officer camera. AB 1034 is being heard in this committee today.
- AB 79 (Weber), would prohibit a peace officer from using deadly force against or intending to injure, intimidate, or disorient a person by utilizing any unmanned, remotely piloted, powered ground of flying equipment except under specified circumstances. AB 79 is pending hearing in this committee.
- AB 793 (Bonta), would provide that a government entity may not seek, from any court, a compulsory process to enforce a reverse-location demand or a reverse-keyword demand, as defined. AB 793 is pending hearing in this committee.
- AB 742 (Jackson), would prohibit the use of canines by peace officers for arrest and apprehension, or in any circumstances to bite a person, but permits their use of canines for search and rescue, explosives detection, and narcotics detection. AB 742 is pending hearing in the Assembly Appropriations Committee.
- AB 1034 (Wilson), would prohibit a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other officer camera. AB 1034 is being heard in this committee today.
- Prior Legislation:
- SB 1038 (Bradford), would have deleted the January 1, 2023 sunset date on provisions of law that prohibit a law enforcement officer from installing, activating or using a biometric surveillance system in connection with a body-worn camera or data collected by a body-worn camera. SB 1038 died on the inactive file in the Senate.
- AB 1281 (Chau), Chapter 268, Statutes of 2020, would require a business in California that uses facial recognition technology to disclose that usage in a physical sign that is clear and conspicuous at the entrance of every location that uses facial recognition technology.
- AB 1215 (Ting), Chapter 579, Statutes of 2019, prohibited a law enforcement officer or agency from installing, activating, or using a biometric surveillance system in connection with a law enforcement agency’s body-worn camera or any other camera.
- SB 21 (Hill), of the 2017-2018 Legislative Session, would have required local law enforcement agencies to have a policy, approved by the local governing body, in place before using surveillance technology. SB 21 was held in the Assembly Appropriations Committee.
- AB 69 (Rodriguez) Chapter 461, Statutes of 2015, requires law enforcement agencies to follow specified best practices when establishing policies and procedures for downloading and storing data from body-worn cameras.
- SB 34 (Hill) Chapter 532, Statutes of 2015, imposed a variety of security, privacy and public hearing requirements on the use of automated license plate recognition systems, as well as a private right of action and provisions for remedies.
- SB 1038 (Bradford), would have deleted the January 1, 2023 sunset date on provisions of law that prohibit a law enforcement officer from installing, activating or using a biometric surveillance system in connection with a body-worn camera or data collected by a body-worn camera. SB 1038 died on the inactive file in the Senate.
REGISTERED SUPPORT / OPPOSITION:
Support
Arcadia Police Officers' Association
Burbank Police Officers' Association
California Coalition of School Safety Professionals
Claremont Police Officers Association
Corona Police Officers Association
Culver City Police Officers' Association
Deputy Sheriffs' Association of Monterey County
Fullerton Police Officers' Association
League of California Cities
Los Angeles School Police Officers Association
Murrieta Police Officers' Association
Newport Beach Police Association
Palos Verdes Police Officers Association
Placer County Deputy Sheriffs' Assocation
Pomona Police Officers' Association
Riverside Police Officers Association
Riverside Sheriffs' Association
Santa Ana Police Officers Association
Upland Police Officers Association
Opposition
ACLU California Action
Asian Americans Advancing Justice - Asian Law Caucus
California Association of Black Lawyers
California Public Defenders Association (CPDA)
Electronic Frontier Foundation
If/when/how: Lawyering for Reproductive Justice
Lawyers Committee for Civil Rights of The San Francisco Bay Area
Lawyers' Committee for Civil Rights of The San Francisco Bay Area
Media Alliance
Mediajustice
San Francisco Public Defender - Racial Justice Committee
Secure Justice
St. James Infirmary
Stop the Musick Coalition
Tenth Amendment Center
Analysis Prepared by:
Discussed in Hearing
Assembly Standing Committee on Privacy and Consumer Protection
Assembly Standing Committee on Public Safety
Bill Author
Bill Co-Author(s):