AB 2027: Worker data: prohibitions: artificial intelligence.
- Session Year: 2025-2026
- House: Assembly
- Latest Version Date: 2026-04-09
Current Status:
In Progress
(2026-05-14: Coauthors revised.)
Introduced
In Committee
Awaiting First Chamber Vote
Passed First Chamber
In Committee
Awaiting Second Chamber Vote
Passed Second Chamber
Enacted
Existing law establishes the Division of Labor Standards Enforcement within the Department of Industrial Relations. Existing law authorizes the division, which is headed by the Labor Commissioner, to enforce the Labor Code and all labor laws of the state, the enforcement of which is not specifically vested in any other officer, board, or commission.
This bill would require an employer, or a vendor acting on behalf of an employer, to collect and process worker data only as strictly necessary to administer the employment relationship and fulfill specific obligations. The bill would prohibit an employer or vendor from using a worker data to train or deploy artificial intelligence to, among other things, replicate, automate, or replace a workers job, and to prohibit an employer or vendor from deploying artificial intelligence trained with worker data prohibit an employer from using a workers personal information, as defined, to train an artificial intelligence system to replicate, automate, or replace a workers job. The bill job, and would prohibit an employer from selling, disclosing, or otherwise providing access to worker data to a third party to train an artificial intelligence system for the purposes of replicating, automating, or replacing a workers personal information to a third party for the purpose of training an artificial intelligence system to replicate, automate, or replace a workers job. The bill would prohibit a vendor providing services to an employer under a contract from providing access to the personal information of an employers worker data to a third party or using an the personal information of an employers worker data to train artificial intelligence, as specified. The bill would require a contract between an employer and vendor to include specific terms, including, among others, a provision that the employer and vendor be jointly liable for a violation of these prohibitions. a requirement that the vendor implement and maintain reasonable security procedures to protect the workers personal information from, among other things, unauthorized or illegal access. The bill would define terms for these provisions, including employer and worker data. personal information.
The bill would require the Labor Commissioner and authorize a public prosecutor to enforce these provisions. The bill would authorize a worker, or their exclusive representative, who suffered a violation of these provisions to bring a civil action for damages, injunctive relief, punitive damages, and attorneys fees and costs. The bill would establish a statutory penalty for a violation of these provisions of up to $500 per employee for each violation.
The bill would include findings that changes proposed by this bill address a matter of statewide concern rather than a municipal affair and, therefore, apply to all cities, including charter cities.
Existing law establishes the Division of Labor Standards Enforcement within the Department of Industrial Relations. Existing law authorizes the division, which is headed by the Labor Commissioner, to enforce the Labor Code and all labor laws of the state, the enforcement of which is not specifically vested in any other officer, board, or commission.
This bill would require an employer, or a vendor acting on behalf of an employer, to collect and process worker data only as strictly necessary to administer the employment relationship and fulfill specific obligations. The bill would prohibit an employer or vendor from using a worker data to train or deploy artificial intelligence to, among other things, replicate, automate, or replace a workers job, and to prohibit an employer or vendor from deploying artificial intelligence trained with worker data to replicate, automate, or replace a workers job. The bill would prohibit an employer from selling, disclosing, or otherwise providing access to worker data to a third party to train an artificial intelligence system for the purposes of replicating, automating, or replacing a workers job. The bill would prohibit a vendor providing services to an employer under a contract from providing access to the employers worker data to a third party or using an employers worker data to train artificial intelligence, as specified. The bill would require a contract between an employer and vendor to include specific terms, including, among others, a provision that the employer and vendor be jointly liable for a violation of these prohibitions. The bill would define terms for these provisions, including employer and worker data.
The bill would require the Labor Commissioner and authorize a public prosecutor to enforce these provisions. The bill would authorize a worker, or their exclusive representative, who suffered a violation of these provisions to bring a civil action for damages, injunctive relief, punitive damages, and attorneys fees and costs. The bill would establish a statutory penalty for a violation of these provisions of up to $500 per employee for each violation.
The bill would include findings that changes proposed by this bill address a matter of statewide concern rather than a municipal affair and, therefore, apply to all cities, including charter cities.
Existing law prohibits an employer from making, adopting, or enforcing any rule, regulation, or policy preventing an employee from disclosing information to a government or law enforcement agency, to a person with authority over the employee, or to another employee who has authority to investigate, discover, or correct the violation or noncompliance, if the employee has reasonable cause to believe that the information discloses a violation of state or federal statute, or a violation of or noncompliance with a local, state, or federal rule or regulation, regardless of whether disclosing the information is part of the employees job duties. Existing law defines specified terms for its purposes, as provided.This bill would make a nonsubstantive change to this provision.Existing law prohibits an employer from making, adopting, or enforcing any rule, regulation, or policy preventing an employee from disclosing information to a government or law enforcement agency, to a person with authority over the employee, or to another employee who has authority to investigate, discover, or correct the violation or noncompliance, if the employee has reasonable cause to believe that the information discloses a violation of state or federal statute, or a violation of or noncompliance with a local, state, or federal rule or regulation, regardless of whether disclosing the information is part of the employees job duties. Existing law defines specified terms for its purposes, as provided.
This bill would make a nonsubstantive change to this provision.
As used in this part:
(a)Artificial intelligence means an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments.
(b)(1)Employer means a person, business, private entity, state, city, county, charter county, city and county, municipality, charter municipality, special district, transit district, the University of California, the California State University, community college district, school district, or any other state or local governmental entity that, directly or indirectly, or through an agent or any other person, employs or exercises control over the wages, benefits, other compensation, hours, working conditions, access to work or job opportunities, or other terms or conditions of employment, of any worker.
(2)Employer includes an employers labor contractor.
(c)Personal information has the same meaning as defined in Section 1798.140 of the Civil Code.
(c)(d)Public prosecutor has the same meaning as defined in Section 180.
(d)(e)Worker means a natural person who is an employee of, or an independent contractor providing service to or through, an employer.
(e)Worker data means any of the following:(1)Information that identifies, relates to, describes, is reasonably capable of being associated with, or could reasonably be linked, directly or indirectly, with the personal information relating to, a worker, regardless of how the information is collected, inferred, or obtained, including, but not limited to, a workers biometric data, employment history, or personal identifying information.(2)Information that in any way relates to or describes, regardless of how the information is collected, inferred, or obtained, how a worker performs their job duties.1571.(a)An employer, or a vendor acting on behalf of an employer, shall collect and process worker data only as strictly necessary to administer the employment relationship and to fulfill specific, affirmatively requested employment-related or legal obligations, and shall not collect or use worker data for any other purpose, including, but not limited to, using worker data to train an artificial intelligence system to replicate, automate, or replace a workers job.(b)1571.(a)An employer, or vendor acting on behalf of an employer, shall not deploy artificial intelligence trained with worker data employer shall not use a workers personal information to train an artificial intelligence system to replicate, automate, or replace a workers job.
(c)(b)An employer shall not sell, disclose, license, transmit, make available, expose, process, monetize, or otherwise provide access to worker data a workers personal information to a third party, including through automated means, embedded technologies, tracking technologies, networked data collection, or transmission mechanisms, to train an artificial intelligence system for the purposes of replicating, automating, or replacing for the purpose of training an artificial intelligence system to replicate, automate, or replace a workers job.
(d)(c)A vendor providing services to an employer under a contract shall not do any of the following:
(1)Transfer, sell, disclose, license, transmit, make available, expose, process, monetize, or otherwise provide access to the personal information of an employers worker data to a third party, including through automated means, embedded technologies, tracking technologies, or networked data collection or transmission mechanisms to train for the purpose of training an artificial intelligence system for the purposes of replicating, automating, or replacing to replicate, automate, or replace a workers job.
(2)Use the personal information of an employers worker data to train any of the vendors artificial intelligence products, sold or leased to other customers, for the purpose of replicating, automating, or replacing a workers job.
(e)(d)A contract between an employer and a vendor for services involving the access or capture of worker data a workers personal information on the employers behalf shall include both of the following terms:
(1)A a requirement that the vendor must implement and maintain reasonable security procedures to protect the worker data workers personal information from unauthorized or illegal access, destruction, use, modification, or disclosure.
(2)A provision that the vendor and employer agree to be jointly and severally liable for a breach of the vendors data security systems or unauthorized access, to the extent the breach or access involves worker data provided by the employer, and for any violation of this part.1572.(a)The Labor Commissioner shall enforce this part, including investigating an alleged violation and ordering appropriate temporary relief to mitigate a violation or maintain the status quo pending the completion of a full investigation or hearing through the procedures set forth in Section 98, 98.3, 98.7, 98.74, or 1197.1, including issuing a citation against an employer who violates this part or filing a civil action. If a citation is issued, the procedures for issuing, contesting, and enforcing judgments for citations and civil penalties issued by the Labor Commissioner shall be the same as those set forth in Section 98.74 or 1197.1, as applicable.
(b)This part may also alternatively be enforced by a public prosecutor pursuant to Chapter 8 (commencing with Section 180) of Division 1.
(c)Alternatively to subdivision (a) or (b), a worker, or their exclusive representative, who has suffered a violation of this part may bring a civil action in a court of competent jurisdiction for damages caused by that adverse action, including punitive damages. violation.
(d)In any civil action brought pursuant to subdivision (a), (b), or (c), the petitioner may seek appropriate temporary or preliminary injunctive relief, punitive damages, and reasonable attorneys fees and costs.
(e)(1)Except as provided in paragraph (2), in addition to any other remedy, an employer who violates this part may be subject to a penalty of up to five hundred dollars ($500) per worker for each violation.
(2)A worker, the Labor Commissioner, or a public prosecutor may recover a penalty under this part as a statutory penalty paid to the worker or a civil penalty, but not both, for the same violation.
(f)An action brought pursuant to this section may be brought in the superior court in any county in which the violation in question is alleged to have occurred or in which the employer resides or transacts business. resides.
(g)This part does not preempt any city, county, or city and county ordinance that provides equal or greater protection to workers who are covered by this part.
SEC. 2.The Legislature finds and declares that Section 1 of this act, adding Part 5.9 (commencing with Section 1570) to Division 2 of the Labor Code, addresses a matter of statewide concern rather than a municipal affair as that term is used in Section 5 of Article XI of the California Constitution. Therefore, Section 1 of this act applies to all cities, including charter cities.
REVISIONS: HeadingLines 2, 3, 4, and 5.As used in this part:
(a)Artificial intelligence means an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments.
(b)(1)Employer means a person, business, private entity, state, city, county, charter county, city and county, municipality, charter municipality, special district, transit district, the University of California, the California State University, community college district, school district, or any other state or local governmental entity that, directly or indirectly, or through an agent or any other person, employs or exercises control over the wages, benefits, other compensation, hours, working conditions, access to work or job opportunities, or other terms or conditions of employment, of any worker.
(2)Employer includes an employers labor contractor.
(c)Public prosecutor has the same meaning as defined in Section 180.
(d)Worker means a natural person who is an employee of, or an independent contractor providing service to or through, an employer.
(e)Worker data means any of the following:
(1)Information that identifies, relates to, describes, is reasonably capable of being associated with, or could reasonably be linked, directly or indirectly, with the personal information relating to, a worker, regardless of how the information is collected, inferred, or obtained, including, but not limited to, a workers biometric data, employment history, or personal identifying information.
(2)Information that in any way relates to or describes, regardless of how the information is collected, inferred, or obtained, how a worker performs their job duties.
1571.(a)An employer, or a vendor acting on behalf of an employer, shall collect and process worker data only as strictly necessary to administer the employment relationship and to fulfill specific, affirmatively requested employment-related or legal obligations, and shall not collect or use worker data for any other purpose, including, but not limited to, using worker data to train an artificial intelligence system to replicate, automate, or replace a workers job.
(b)An employer, or vendor acting on behalf of an employer, shall not deploy artificial intelligence trained with worker data to replicate, automate, or replace a workers job.
(c)An employer shall not sell, disclose, license, transmit, make available, expose, process, monetize, or otherwise provide access to worker data to a third party, including through automated means, embedded technologies, tracking technologies, networked data collection, or transmission mechanisms, to train an artificial intelligence system for the purposes of replicating, automating, or replacing a workers job.
(d)A vendor providing services to an employer under a contract shall not do any of the following:
(1)Transfer, sell, disclose, license, transmit, make available, expose, process, monetize, or otherwise provide access to the employers worker data to a third party, including through automated means, embedded technologies, tracking technologies, or networked data collection or transmission mechanisms to train an artificial intelligence system for the purposes of replicating, automating, or replacing a workers job.
(2)Use the employers worker data to train any of the vendors artificial intelligence products, sold or leased to other customers, for the purpose of replicating, automating, or replacing a workers job.
(e)A contract between an employer and a vendor for services involving the access or capture of worker data on the employers behalf shall include both of the following terms:
(1)A requirement that the vendor must implement and maintain reasonable security procedures to protect the worker data from unauthorized or illegal access, destruction, use, modification, or disclosure.
(2)A provision that the vendor and employer agree to be jointly and severally liable for a breach of the vendors data security systems or unauthorized access, to the extent the breach or access involves worker data provided by the employer, and for any violation of this part.
1572.(a)The Labor Commissioner shall enforce this part, including investigating an alleged violation and ordering appropriate temporary relief to mitigate a violation or maintain the status quo pending the completion of a full investigation or hearing through the procedures set forth in Section 98, 98.3, 98.7, 98.74, or 1197.1, including issuing a citation against an employer who violates this part or filing a civil action. If a citation is issued, the procedures for issuing, contesting, and enforcing judgments for citations and civil penalties issued by the Labor Commissioner shall be the same as those set forth in Section 98.74 or 1197.1, as applicable.
(b)This part may also alternatively be enforced by a public prosecutor pursuant to Chapter 8 (commencing with Section 180) of Division 1.
(c)Alternatively to subdivision (a) or (b), a worker, or their exclusive representative, who has suffered a violation of this part may bring a civil action in a court of competent jurisdiction for damages caused by that adverse action, including punitive damages.
(d)In any civil action brought pursuant to subdivision (a), (b), or (c), the petitioner may seek appropriate temporary or preliminary injunctive relief, punitive damages, and reasonable attorneys fees and costs.
(e)(1)Except as provided in paragraph (2), in addition to any other remedy, an employer who violates this part may be subject to a penalty of up to five hundred dollars ($500) per worker for each violation.
(2)A worker, the Labor Commissioner, or a public prosecutor may recover a penalty under this part as a statutory penalty paid to the worker or a civil penalty, but not both, for the same violation.
(f)An action brought pursuant to this section may be brought in the superior court in any county in which the violation in question is alleged to have occurred or in which the employer resides or transacts business.
(g)This part does not preempt any city, county, or city and county ordinance that provides equal or greater protection to workers who are covered by this part.
SEC. 2.The Legislature finds and declares that Section 1 of this act, adding Part 5.9 (commencing with Section 1570) to Division 2 of the Labor Code, addresses a matter of statewide concern rather than a municipal affair as that term is used in Section 5 of Article XI of the California Constitution. Therefore, Section 1 of this act applies to all cities, including charter cities.
SECTION 1.Section 1102.5 of the Labor Code is amended to read:1102.5.(a)An employer, or any person acting on behalf of the employer, shall not make, adopt, or enforce any rule, regulation, or policy preventing an employee from disclosing information to a government or law enforcement agency, to a person with authority over the employee, or to another employee who has authority to investigate, discover, or correct the violation or noncompliance, or from providing information to, or testifying before, any public body conducting an investigation, hearing, or inquiry, if the employee has reasonable cause to believe that the information discloses a violation of state or federal statute, or a violation of or noncompliance with a local, state, or federal rule or regulation, regardless of whether disclosing the information is part of the employees job duties.(b)An employer, or any person acting on behalf of the employer, shall not retaliate against an employee for disclosing information, or because the employer believes that the employee disclosed or may disclose information, to a government or law enforcement agency, to a person with authority over the employee or another employee who has the authority to investigate, discover, or correct the violation or noncompliance, or for providing information to, or testifying before, any public body conducting an investigation, hearing, or inquiry, if the employee has reasonable cause to believe that the information discloses a violation of state or federal statute, or a violation of or noncompliance with a local, state, or federal rule or regulation, regardless of whether disclosing the information is part of the employees job duties.(c)An employer, or any person acting on behalf of the employer, shall not retaliate against an employee for refusing to participate in an activity that would result in a violation of state or federal statute, or a violation of or noncompliance with a local, state, or federal rule or regulation.(d)An employer, or any person acting on behalf of the employer, shall not retaliate against an employee for having exercised their rights under subdivision (a), (b), or (c) in any former employment.(e)A report made by an employee of a government agency to their employer is a disclosure of information to a government or law enforcement agency pursuant to subdivisions (a) and (b).(f)(1)In addition to other remedies available, an employer is liable for a civil penalty not exceeding ten thousand dollars ($10,000) per employee for each violation of this section to be awarded to the employee who was retaliated against.(2)In assessing this penalty, the Labor Commissioner shall consider the nature and seriousness of the violation based on the evidence obtained during the course of the investigation. The Labor Commissioners consideration of the nature and seriousness of the violation shall include, but is not limited to, the type of violation, the economic or mental harm suffered, and the chilling effect on the exercise of employment rights in the workplace, and shall be considered to the extent evidence obtained during the investigation concerned any of these or other relevant factors.(g)This section does not apply to rules, regulations, or policies that implement, or to actions by employers against employees who violate, the confidentiality of the lawyer-client privilege of Article 3 (commencing with Section 950) of, or the physician-patient privilege of Article 6 (commencing with Section 990) of, Chapter 4 of Division 8 of the Evidence Code, or trade secret information.(h)An employer, or a person acting on behalf of the employer, shall not retaliate against an employee because the employee is a family member of a person who has, or is perceived to have, engaged in any acts protected by this section.(i)For the purposes of this section, employer or a person acting on behalf of the employer includes, but is not limited to, a client employer as defined in paragraph (1) of subdivision (a) of Section 2810.3 and an employer listed in subdivision (b) of Section 6400.(j)The court is authorized to award reasonable attorneys fees to a plaintiff who brings a successful action for a violation of these provisions.(a)An employer, or any person acting on behalf of the employer, shall not make, adopt, or enforce any rule, regulation, or policy preventing an employee from disclosing information to a government or law enforcement agency, to a person with authority over the employee, or to another employee who has authority to investigate, discover, or correct the violation or noncompliance, or from providing information to, or testifying before, any public body conducting an investigation, hearing, or inquiry, if the employee has reasonable cause to believe that the information discloses a violation of state or federal statute, or a violation of or noncompliance with a local, state, or federal rule or regulation, regardless of whether disclosing the information is part of the employees job duties.
(b)An employer, or any person acting on behalf of the employer, shall not retaliate against an employee for disclosing information, or because the employer believes that the employee disclosed or may disclose information, to a government or law enforcement agency, to a person with authority over the employee or another employee who has the authority to investigate, discover, or correct the violation or noncompliance, or for providing information to, or testifying before, any public body conducting an investigation, hearing, or inquiry, if the employee has reasonable cause to believe that the information discloses a violation of state or federal statute, or a violation of or noncompliance with a local, state, or federal rule or regulation, regardless of whether disclosing the information is part of the employees job duties.
(c)An employer, or any person acting on behalf of the employer, shall not retaliate against an employee for refusing to participate in an activity that would result in a violation of state or federal statute, or a violation of or noncompliance with a local, state, or federal rule or regulation.
(d)An employer, or any person acting on behalf of the employer, shall not retaliate against an employee for having exercised their rights under subdivision (a), (b), or (c) in any former employment.
(e)A report made by an employee of a government agency to their employer is a disclosure of information to a government or law enforcement agency pursuant to subdivisions (a) and (b).
(f)(1)In addition to other remedies available, an employer is liable for a civil penalty not exceeding ten thousand dollars ($10,000) per employee for each violation of this section to be awarded to the employee who was retaliated against.
(2)In assessing this penalty, the Labor Commissioner shall consider the nature and seriousness of the violation based on the evidence obtained during the course of the investigation. The Labor Commissioners consideration of the nature and seriousness of the violation shall include, but is not limited to, the type of violation, the economic or mental harm suffered, and the chilling effect on the exercise of employment rights in the workplace, and shall be considered to the extent evidence obtained during the investigation concerned any of these or other relevant factors.
(g)This section does not apply to rules, regulations, or policies that implement, or to actions by employers against employees who violate, the confidentiality of the lawyer-client privilege of Article 3 (commencing with Section 950) of, or the physician-patient privilege of Article 6 (commencing with Section 990) of, Chapter 4 of Division 8 of the Evidence Code, or trade secret information.
(h)An employer, or a person acting on behalf of the employer, shall not retaliate against an employee because the employee is a family member of a person who has, or is perceived to have, engaged in any acts protected by this section.
(i)For the purposes of this section, employer or a person acting on behalf of the employer includes, but is not limited to, a client employer as defined in paragraph (1) of subdivision (a) of Section 2810.3 and an employer listed in subdivision (b) of Section 6400.
(j)The court is authorized to award reasonable attorneys fees to a plaintiff who brings a successful action for a violation of these provisions.
|
Date
|
Status
|
|---|---|
| May 14, 2026 | Coauthors revised. |
| May 14, 2026 | In committee: Held under submission. |
| May 6, 2026 | In committee: Set, first hearing. Referred to APPR. suspense file. |
| Apr 22, 2026 | From committee: Do pass and re-refer to Com. on APPR. (Ayes 10. Noes 3.) (April 21). Re-referred to Com. on APPR. |
| Apr 16, 2026 | In committee: Set, first hearing. Hearing canceled at the request of author. |
| Apr 13, 2026 | Re-referred to Com. on P. & C.P. |
| Apr 9, 2026 | From committee chair, with author's amendments: Amend, and re-refer to Com. on P. & C.P. Read second time and amended. |
| Apr 9, 2026 | From committee: Do pass and re-refer to Com. on P. & C.P. (Ayes 5. Noes 0.) (April 8). Re-referred to Com. on P. & C.P. |
| Mar 17, 2026 | Re-referred to Com. on L. & E. |
| Mar 16, 2026 | From committee chair, with author's amendments: Amend, and re-refer to Com. on L. & E. Read second time and amended. |
| Mar 16, 2026 | Referred to Coms. on L. & E., P. & C.P. and JUD. |
| Feb 18, 2026 | From printer. May be heard in committee March 20. |
| Feb 17, 2026 | Read first time. To print. |
|
Date
|
Description
|
Location
|
Votes
|
|
|---|---|---|---|---|
| Apr 21, 2026 | Do pass and be re-referred to the Committee on [Appropriations] | Assembly Standing Committee on Privacy and Consumer Protection |
Passed |
|
| Apr 8, 2026 | Do pass and be re-referred to the Committee on [Privacy and Consumer Protection] | Assembly Standing Committee on Labor and Employment |
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
|
|---|---|---|
| Acclamation Insurance Management Services | ** Against | |
| Alameda Labor Council | ** Support | |
| Allied Managed Care | ** Against | |
| American Federation of Musicians, Local 7 | ** Support | |
| American Federation of State, County and Municipal Employees (AFSCME) California | ** Support | |
| Associated General Contractors Of California | ** Against | |
| Associated General Contractors San Diego | ** Against | |
| Association of California Healthcare Districts | ** Against | |
| Association Of California Healthcare Districts (Achd) | ** Against | |
| Association Of California School Administrators | ** Against | |
| California Alliance for Retired Americans (CARA) | ** Support | |
| California Apartment Association | ** Against | |
| California Chamber Of Commerce | ** Against | |
| California Faculty Association | ** Support | |
| California Farm Bureau | ** Against | |
| California Federation of Labor Unions | ** Support | |
| California Federation of Teachers Afl-cio | ** Support | |
| California Hospital Association | ** Against | |
| California Manufactures & Technology Association | ** Against | |
| California Retailers Association | ** Against | |
| California School Employees Association | ** Support | |
| California Special Districts Association | ** Against | |
| California State Association of Counties (CSAC) | ** Against | |
| California State Legislative Board of the Sheet Metal, Air, Rail and Transportation Workers - Transportation Division (SMART-TD) | ** Support | |
| California Teachers Association | ** Support | |
| California Trucking Association | ** Against | |
| Central Coast Labor Council | ** Support | |
| Central Labor Council, Fresno-Madera-Tulare-Kings Counties, AFL-CIO | ** Support | |
| Cinema Association of California | ** Against | |
| Civil Justice Association of California (CJAC) | ** Against | |
| Flasher Barricade Association | ** Against | |
| Inland Empire Labor Council, AFL-CIO | ** Support | |
| Insights Association | ** Against | |
| LeadingAge California | ** Against | |
| League Of California Cities | ** Against | |
| Los Angeles Area Chamber Of Commerce | ** Against | |
| National Association Of Professional Employer Organizations | ** Against | |
| National Payroll Reporting Consortium | ** Against | |
| North Bay Labor Council | ** Support | |
| North Valley Labor Federation | ** Support | |
| Oakland Privacy | ** Support | |
| Orange County Labor Federation, AFL-CIO | ** Support | |
| Rural County Representatives of California (RCRC) | ** Against | |
| san mateo county central labor council | ** Support | |
| Techequity Action | ** Support | |
| Technet | ** Against | |
| UAW Region 6 | ** Support | |
| Urban Counties of California (UCC) | ** Against | |
| What We Will | ** Support |
Date of Hearing: May 6, 2026
ASSEMBLY COMMITTEE ON APPROPRIATIONS
Buffy Wicks, Chair
ABPCA Bill Id:AB 2027
Author:(Ward) – As Amended Ver:April 9, 2026
Policy Committee: | Labor and Employment | Vote: | 5 - 0 |
Privacy and Consumer Protection | 10 - 3 | ||
|
|
Urgency: No State Mandated Local Program: No Reimbursable: No
SUMMARY:
This bill prohibits an employer from using a worker’s personal information to train an artificial intelligence (AI) system to replicate, automate, or replace a worker’s job.
Specifically, this bill:
- Prohibits an employer from using a worker’s personal information to train an AI system to replicate, automate, or replace a worker’s job and prohibits an employer from selling, disclosing, or otherwise providing access to a worker’s personal information to a third party, including through embedded technologies, for the purpose of training an AI system to replicate, automate, or replace a worker’s job.
- Prohibits a vendor providing services to an employer under contract from providing access to a worker’s personal information to a third party for the purpose of training an AI system, prohibits a vendor from using a worker’s personal information to train the vendor’s AI products, and requires a contract between an employer and a vendor to include a requirement that the vendor implement and maintain reasonable security procedures to protect the worker’s personal information from unauthorized or illegal access.
- Requires the Labor Commissioner (LC) to enforce these worker personal information provisions, including issuing a citation and filing a civil action, and authorizes enforcement by a public prosecutor, worker, or worker’s representative through a civil action.
- Allows the petitioner in a civil action to seek temporary or preliminary injunctive relief, punitive damages, and attorney’s fees and costs, and allows an employer in violation to be subject to a penalty of up to $500 per violation.
FISCAL EFFECT:
- Costs of an unknown, but potentially significant amount, in excess of $150,000, to the state as an employer to ensure compliance with this bill’s worker personal information provisions (General Fund (GF) or special fund). The magnitude of costs depends the state’s existing data use policies, especially in relation to vendors under contract.
Similarly, potentially significant costs to local government employers, which the state would not need to reimburse because this bill’s new requirements for local agencies equally apply to the private sector.
- Costs of an unknown, but likely significant amount, potentially in the millions of dollars, to the LC to enforce this bill’s worker personal information provisions (Labor Enforcement and Compliance Fund). The LC’s primary mission is enforcement of the state’s wage and hour laws and protecting employees from related retaliation, whereas this bill regulates employer conduct that generally is outside the usual scope of enforcement for the LC. For example, an employer that utilizes scheduling software may be subject to LC enforcement if the program developer uses that employer’s worker’s personal information for unbeknownst purposes in violation of this bill.
- The Department of Justice (DOJ) reports no significant impact from this bill, indicating DOJ is unlikely to experience an increase in litigation representing state agencies for alleged violations of this bill’s provisions or pursue civil actions to enforce this bill’s provisions as a public prosecutor.
- Ongoing cost pressures of an unknown amount to the courts in additional workload by authorizing a new civil action against an employer that violates this bill’s worker personal information provisions (GF or Trial Court Trust Fund (TCTF)). It is unclear how many civil actions may be filed statewide and how much court time may be needed to resolve each case, but it generally costs approximately $1,000 to operate a courtroom for one hour. Although courts are not funded on the basis of workload, increased pressure on staff and the TCTF may create a demand for increased court funding from the GF. The state budget provides annual GF backfills to the TCTF to offset revenue reductions, totaling approximately $117.3 million in fiscal year 2025-26.
The Legislative Analyst’s Office recently warned of GF structural deficits of around $35 billion per year in the 2027-28 fiscal year and ongoing.
COMMENTS:
- Purpose. According to the author:
AB 2027 sets clear boundaries. It prohibits employers from using worker data to train AI systems that automate or replace jobs, and prevents the sale or sharing of that data for that purpose. It also ensures tech developers cannot use worker data collected through their tools to build systems that displace workers.
This bill affirms a simple principle: workers should not be used to train their own replacements, and innovation must not come at the expense of economic security.
- Workplace AI Tools. While employers surveilling their workers, both during and after work hours, is far from a new phenomenon, advances in affordable surveillance and AI technology has made that surveillance much more intrusive. Many workers, while generally aware they are being monitored, are not aware of the extent of the surveillance or what is being done with the information. Technology such as motion scanners, Bluetooth beacons, and keystroke logging track every movement of workers in the office to gauge productivity, with some workplaces even using biometric data, such as eye movements, body shifts, and facial expressions captured via webcam, to evaluate whether workers are attentive in their duties. An employer may then use data gathered from workplace AI tools to make a wide range of decisions affecting a worker’s wages and benefits, work schedule, performance evaluation, discipline, promotion, and other workplace conditions. The data may also be aggregated, labeled, and repurposed as training data for AI systems designed to optimize workflows, automate tasks, and in some cases, replace the very workers from whom the data was derived.
This bill prohibits an employer, or a vendor that contracts with an employer, from using a worker’s personal information to train an AI system to replicate, automate, or replace a worker’s job, either directly or by providing a third party access to the data. This bill also requires a contract between an employer and a vendor to contain reasonable security measures to protect a worker’s personal information.
- Support and Opposition. This bill is sponsored by the California Federation of Labor Unions, which argues this bill “is essential in the fight to protect workers from being used to train their own replacements.” This bill is also supported by a large coalition of labor organizations, as well as privacy rights groups.
This bill is opposed by a large coalition of business groups, led by the California Chamber of Commerce, which argues this bill “would impose sweeping and unworkable restrictions on the use of worker data that will significantly hinder responsible innovation and expose employers to substantial litigation risk.”
- Related Legislation. AB 1221 (Bryan) regulates an employer’s use of workplace surveillance tools and worker data. AB 1221 was held on this committee’s suspense file.
AB 1331 (Elhawary) prohibits an employer from using a workplace surveillance tool to monitor workers in employee-only areas and during off-duty hours. AB 1331 was ordered to the Senate Inactive File.
AB 1883 (Bryan) prohibits an employer from using workplace surveillance tools for certain purposes. AB 1883 is pending on this committee’s suspense file.
AB 1898 (Schultz) requires an employer to notify a worker of the employer’s use of a workplace AI tool. AB 1898 is pending on this committee’s suspense file.
SB 7 (McNerney) would have required an employer to provide written notice to workers regarding the use of certain automated decision systems making employment-related decisions. SB 7 was vetoed by Governor Newsom, who stated, “rather than addressing the specific ways employers misuse this technology, the bill imposes unfocused notification requirements on any business using even the most innocuous tools. This proposed solution fails to directly address incidents of misuse.”
SB 238 (Smallwood-Cuevas) requires an employer to annually provide notice of all the workplace surveillance tools used in the workplace to the Department of Industrial Relations. SB 238 is pending hearing by the Assembly Privacy and Consumer Protection Committee.
SB 947 (McNerney) prohibits an employer from using an automated decision system (ADS) to perform certain functions and regulates allowable ADS uses. SB 947 is pending on the Senate Appropriations Committee’s suspense file.
SB 951 (Reyes) requires an employer to provide 90-day written notice to workers and certain entities before a technological displacement affecting 25 or more workers or 25% of the workforce, whichever is less. SB 951 is pending on the Senate Appropriations Committee’s suspense file.
Analysis Prepared by: Irene Ho / APPR. / (916) 319-2081
Date of Hearing: April 21, 2026
Fiscal: Yes
ASSEMBLY COMMITTEE ON PRIVACY AND CONSUMER PROTECTION
Rebecca Bauer-Kahan, Chair
ABPCA Bill Id:AB 2027 (
Author:Ward) – As Amended Ver:April 9, 2026
SUBJECT: Worker data: prohibitions: artificial intelligence
SYNOPSIS
As surveillance technologies have become cheaper, more scalable, and easier to deploy, worker privacy has steadily eroded across a wide range of industries. Employers can now monitor keystrokes, employee location, communications, productivity metrics, and even biometric signals with relative ease, even extending these practices to contractors and third-party workers. Once collected, this workplace data does not go unused – it can be aggregated, labeled, and repurposed as training data for artificial intelligence (AI) systems designed to optimize workflows, automate tasks, and, in some cases, replace the very workers from whom the data was derived.
This bill would prohibit an employer, or a third-party vendor that contracts with an employer, from using a worker’s personal information to train an AI system to replicate, automate, or replace the worker’s job – either directly, or by providing a third party with access to the data. The bill would additionally require contracts between employers and third-party vendors to contain reasonable security measures to protect workers’ personal information. The bill is enforceable by the Labor Commissioner, public prosecutors, or through a civil action brought by a worker or their exclusive representative.
AB 2027 is sponsored by the California Federation of Labor Unions and supported by a variety of workers’ rights and privacy advocacy groups. It is opposed by a collection of trade associations including the California Chamber of Commerce.
EXISTING LAW:
- Provides, pursuant to the California Constitution, that all people are free and independent by nature and have inalienable rights. Among these is the fundamental right to privacy. (Cal. Const. art. I, § 1.)
- Establishes the Information Practices Act of 1977, governing the collection, maintenance, use, and dissemination of personal information by California state agencies. (Civ. Code § 1798 et seq.)
- Provides the following definition for “personal information”:
(1) “Personal information” means information that identifies, relates to, describes, is reasonably capable of being associated with, or could reasonably be linked, directly or indirectly, with a particular consumer or household. Personal information includes, but is not limited to, the following if it identifies, relates to, describes, is reasonably capable of being associated with, or could be reasonably linked, directly or indirectly, with a particular consumer or household:
(A) Identifiers such as a real name, alias, postal address, unique personal identifier, online identifier, Internet Protocol address, email address, account name, social security number, driver’s license number, passport number, or other similar identifiers.
(B) Any personal information described in subdivision (e) of Section 1798.80.
(C) Characteristics of protected classifications under California or federal law.
(D) Commercial information, including records of personal property, products or services purchased, obtained, or considered, or other purchasing or consuming histories or tendencies.
(E) Biometric information.
(F) Internet or other electronic network activity information, including, but not limited to, browsing history, search history, and information regarding a consumer’s interaction with an internet website application, or advertisement.
(G) Geolocation data.
(H) Audio, electronic, visual, thermal, olfactory, or similar information.
(I) Professional or employment-related information.
(J) Education information, defined as information that is not publicly available personally identifiable information as defined in the Family Educational Rights and Privacy Act (20 U.S.C. Sec. 1232g; 34 C.F.R. Part 99).
(K) Inferences drawn from any of the information identified in this subdivision to create a profile about a consumer reflecting the consumer’s preferences, characteristics, psychological trends, predispositions, behavior, attitudes, intelligence, abilities, and aptitudes.
(L) Sensitive personal information.
(2) (A) “Personal information” does not include publicly available information or lawfully obtained, truthful information that is a matter of public concern.
(B) (i) For purposes of this paragraph, “publicly available” means any of the following:
(I) Information that is lawfully made available from federal, state, or local government records.
(II) Information that a business has a reasonable basis to believe is lawfully made available to the general public by the consumer or from widely distributed media.
(III) Information made available by a person to whom the consumer has disclosed the information if the consumer has not restricted the information to a specific audience.
(ii) “Publicly available” does not mean biometric information collected by a business about a consumer without the consumer’s knowledge.
(3) “Personal information” does not include consumer information that is deidentified or aggregate consumer information.
(4) “Personal information” can exist in various formats, including, but not limited to, all of the following:
(A) Physical formats, including paper documents, printed images, vinyl records, or video tapes.
(B) Digital formats, including text, image, audio, or video files.
(C) Abstract digital formats, including compressed or encrypted files, metadata, or artificial intelligence systems that are capable of outputting personal information. (Civ. Code § 1798.140.)
- Establishes the California Consumer Privacy Act, which grants consumers certain rights regarding their personal information, including enhanced notice, access, and disclosure; the right to deletion; the right to restrict the sale of information; and protection from discrimination for exercising these rights. (Civ. Code § 1798.100 et seq.)
- Establishes the Division of Labor Standards Enforcement (DLSE), under the direction of the Labor Commissioner, within the Department of Industrial Relations, and authorizes the LC to investigate employee complaints and enforce labor laws, as specified. (Lab. Code § 79 et seq.)
- Authorizes the DLSE to enforce the Labor Code and all labor laws of the state, the enforcement of which is not specifically vested in any other officer, board, or commission. (Lab. Code § 95(a).)
- Authorizes the Labor Commissioner to prosecute actions for the collection of wages, penalties, and demands, as specified. (Lab. Code § 98.3.)
- Authorizes, until January 1, 2029, the Attorney General, a district attorney, a city attorney, a county counsel, or any other city or county prosecutor, to prosecute an action, either civil or criminal, for violation of certain provisions of the Labor Code or to enforce those provisions independently. (Lab. Code § 180 et seq.)
THIS BILL:
- Defines:
- “Artificial intelligence” to mean an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments.
- “Employer” to mean a person, business, private entity, state, city, county, charter county, city and county, municipality, charter municipality, special district, transit district, the University of California, the California State University, community college district, school district, or any other state or local governmental entity that, directly or indirectly, or through an agent or any other person, employs or exercises control over the wages, benefits, other compensation, hours, working conditions, access to work or job opportunities, or other terms or conditions of employment, of any worker. Further defines “employer” to include an employer’s labor contractor.
- “Personal information” to have the same meaning as defined in Section 1798.140 of the Civil Code.
- “Public prosecutor” to have the same meaning as defined in Section 180 of the Labor Code.
- “Worker” to mean a natural person who is an employee of, or an independent contractor providing service to or through, an employer.
- Prohibits an employer from using a worker’s personal information to train an artificial intelligence system to replicate, automate, or replace a worker’s job.
- Prohibits an employer from selling, disclosing, licensing, transmitting, making available, exposing, processing, monetizing, or otherwise providing access to a worker’s personal information to a third party for the purpose of training an artificial intelligence system to replicate, automate, or replace a worker’s job.
- Prohibits a vendor from providing services to an employer under a contract from doing any of the following:
- Transferring, selling, disclosing, licensing, transmitting, making available, exposing, processing, monetizing, or otherwise providing access to the personal information of an employer’s worker to a third party for the purpose of training an artificial intelligence system to replicate, automate, or replace a worker’s job.
- Using the personal information of an employer’s worker to train an artificial intelligence system to replicate, automate, or replace a worker’s job if the vendor sells or leases the artificial intelligence system to other customers.
- Requires that a contract between an employer and a vendor for services involving the access or capture of a worker’s personal information on the employer’s behalf include a requirement that the vendor must implement and maintain reasonable security procedures to protect the worker’s personal information from unauthorized or illegal access, destruction, use, modification, or disclosure.
- Provides that the Labor Commissioner or a public prosecutor shall enforce the provisions of the bill, or that a worker or their exclusive representative who suffers a violation may bring a civil action, and authorizes a penalty of up to $500 for each violation in addition to any injunctive relief, punitive damages, and reasonable attorney’s fees and costs authorized by the court. Specifies that actions may be brought in the superior court in any county in which the violation in question is alleged to have occurred, or in which the employer resides. Specifies that enforcement provisions do not preempt any city, county, or city and county ordinance that provides equal or greater protection to workers.
COMMENTS:
- Author’s statement. According to the author:
Artificial intelligence is rapidly reshaping our economy, but it is increasingly being built on the backs of workers without their knowledge or consent. The data workers generate every day is fueling AI systems that can monitor, analyze, and ultimately replace them, while employers and tech companies profit from its extraction and sale.
While workers have historically faced being asked to train their replacements, the scale and speed of today’s technology is unprecedented. AI systems can now learn from human labor in real time, accelerating job loss and deskilling across industries, from call centers to software engineering. This is not just about privacy or dignity, it is a direct threat to jobs and economic stability.
AB 2027 sets clear boundaries. It prohibits employers from using worker data to train AI systems that automate or replace jobs, and prevents the sale or sharing of that data for that purpose. It also ensures tech developers cannot use worker data collected through their tools to build systems that displace workers.
This bill affirms a simple principle: workers should not be used to train their own replacements, and innovation must not come at the expense of economic security.
- Background. Artificial intelligence. AI refers to the mimicking of human intelligence by artificial systems such as computers.[1] AI uses algorithms – sets of rules – to transform inputs into outputs. Inputs and outputs can be anything a computer is able to process: numbers, text, audio, or video, for example. AI is not fundamentally different from other computer functions; unlike other computer functions, however, AI is able to accomplish tasks that are normally performed by humans.
Most modern AI tools are created through a process known as “machine learning,” which enables AI tools to learn relationships between inputs and outputs without being explicitly programmed.[2] The process of exposing a naïve AI to data is known as “training.” The algorithm that an AI develops during training is known as its “model.” At its core, training is an optimization problem: machine learning attempts to identify model parameters – weights – that minimize the difference between predicted outcomes and actual outcomes. During training, these weights are continuously adjusted to improve the model’s performance. Once trained, the model can process new, never-before-seen data.[3]
Models trained on small, specific datasets in order to make recommendations and predictions are sometimes referred to as “predictive AI.” This differentiates them from generative AI (GenAI) which are trained on massive datasets in order to produce detailed text, images, audio, and video. When ChatGPT generates text in clear, concise paragraphs, it uses GenAI that has been trained on the written contents of the internet.[4] When Netflix suggests content to a viewer, its recommendation is produced using predictive AI that has been trained on the viewing habits of Netflix users.[5]
AI’s effect on labor markets. According to a recent report from executive outplacement firm Challenger, Gray & Christmas, AI was responsible for 54,836 announced layoffs in the United States in 2025.[6] Earlier this year, a Resume.org survey of 1000 United States hiring managers revealed that 55% of companies expected further layoffs in 2026, and that 44% of these companies cited AI as a top driver – though 59% admitted they intentionally emphasized AI when explaining hiring freezes or layoffs because it “plays better with stakeholders.” [7] It is difficult to separate the signal from the noise when assessing the effect that advanced AI is having on the labor market, as emphasized in a recent review published in Computers in Human Behavior Reports:
The use of Generative AI has significant effects on businesses and the labor market. One of the key contributions of this technology is the ability to generate content. However, the generation of inaccurate and incomplete content by AI may lead to a decrease in trust in the technology. In some industries and occupations, Generative AI may replace employees in content creation activities and related tasks, reducing job opportunities in these sectors. On the other hand, GenAI may create new specializations in content creation, leading to new job opportunities for people with AI-related knowledge.[8]
Ford CEO Jim Farley recently expressed his belief that “artificial intelligence is going to replace literally half of all white collar workers,”[9] a belief shared by Anthropic CEO Darius Amodei.[10] On the other hand, Nvidia CEO Jensen Huang recently predicted that AI will primarily augment, rather than replace, American workers.[11] Anthropic’s “Economic Index” tool, which tracks AI usage trends across the workforce, reveals that California currently ranks fifth among states in terms of AI usage, behind Washington DC, Massachusetts, Washington state, and New York.[12]
California’s creative economy. The creative economy is both hugely important to California and particularly sensitive to AI. The Otis College 2025 Report on the Creative Economy estimates that California’s creative economy employs more than 760,000 workers with an average salary of $191,000 per year, accounting for roughly 5% of the state’s workforce.[13] The 2023 version of this report estimated that California’s creative economy was responsible for 14.9% of the state’s $3.4 trillion economy in 2021, generating over $194.1 billion in federal, state, and municipal taxes.[14]
While the adoption of GenAI is opening up new possibilities in music, film, publishing and other creative fields, the technology’s impact on California’s creative economy and individual creators remains uncertain and unevenly distributed. To date, few studies have assessed how AI will affect jobs and incomes for writers, actors, musicians, designers and other creators. A global report published by the International Confederation of Societies of Authors and Composers (CISAC and PMP Strategy) in December of 2024 projected the emergence of a massive market for AI-generated content. The study estimated that AI-generated audiovisual content, including film and TV, could be worth €48 billion ($55.73 billion) by 2028. However, the report warned that up to 21% of creators’ revenue in film and TV could be cannibalized as a result:
In an unchanged regulatory framework, creators will actually suffer losses on two fronts: the loss of revenues due to the unauthorised use of their works by Gen AI models without remuneration; and replacement of their traditional revenue streams due to the substitution effect of AI-generated outputs, competing against human- made works.[15]
Tech industry layoffs. Over the past two years, major California-based tech companies have laid off hundreds of thousands of workers around the globe. These companies have increasingly linked these layoffs to AI: in March, Reuters reported that Meta’s layoff plans, which could affect 20% or more of the company, were intended to offset AI infrastructure investments and prepare for greater efficiency brought about by AI-assisted workers.[16] Oracle similarly laid off 30,000 employees in late March – representing roughly 18% of its global workforce – in order to fund its commitment to aggressively build out AI infrastructure.[17] Amazon announced plans to lay off 16,000 employees in January of this year, its second round of large-scale job reductions in three months. The company announced in a blog post that the layoffs were necessary to increase its decision-making speed.[18] Jack Dorsey, founder and CEO of startup “Block, Inc” tied cutting 40% of his company to advances in efficiency due to AI in a recent X post:
we're not making this decision because we're in trouble. our business is strong. gross profit continues to grow, we continue to serve more and more customers, and profitability is improving. but something has changed. we're already seeing that the intelligence tools we’re creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company. and that's accelerating rapidly.[19]
Workplace surveillance. Over the last five years, surveillance tools have become more affordable and more intrusive. Employers can collect vast dossiers on their employees, gathering sweeping amounts of data about every aspect of their jobs and personal lives. Often this is done “without employees’ full informed or free consent. Many workers, while generally aware they are being monitored, don’t know the extent of the surveillance or what is being done with the information.”[20]
Employers are using more surveillance technology than ever — digital cameras, motion scanners, RFID badges, Apple Watch badges, Bluetooth beacons, keystroke logging — to track
every single movement of workers in the office and to gauge their productivity. Some workplaces are using biometric data such as eye movements, body shifts, and facial expressions,
captured by computer webcams, to evaluate whether their employees are being appropriately
attentive in their work tasks. As an example, AI systems at call centers record and grade how workers are handling calls. This technology can be used to “coach” workers while they are talking to customers, telling them to sound happier or be more empathetic.[21] Another example is wearable technology that, among other things, tracks a worker’s movements throughout the day, gathering biometric data, measuring how many times they use the bathroom, how long they spend in break areas, and which employees are spending time together. At least one company sells biometric ID badges with microphones, sensors, and other tools to record conversations, monitor speech, body movements, and location.[22] Even body temperature, sweat, and frequency of bathroom visits can be tracked and analyzed by employers.
A recent article in MIT Technology Review describes one company’s surveillance tool this way:
Companies that use electronic employee monitoring report that they are most often looking to the technologies not only to increase productivity but also to manage risk. And software like Teramind[23] offers tools and analysis to help with both priorities. While Teramind, a globally distributed company, keeps its list of over 10,000 client companies private, it provides resources for the financial, health-care, and customer service industries, among others—some of which have strict compliance requirements that can be tricky to keep on top of. The platform allows clients to set data-driven standards for productivity, establish thresholds for alerts about toxic communication tone or language, create tracking systems for sensitive file sharing, and more.
[. . .]
Selecting and tuning the appropriate combination of data is up to Teramind’s clients and depends on the size, goals, and capabilities of the particular company. The companies are also the ones to decide, based on their legal and compliance requirements, what measures to take if thresholds for negative behavior or low performance are hit.[24]
According to documents filed by Amazon workers with the National Labor Relations Board, Amazon tracks every minute that their workers spend not performing their assigned tasks. To facilitate this, Amazon uses handheld scanners that are also used to track packages. One worker claims they “can receive a written warning for accumulating 30 minutes of time off task in a day one time in a rolling one-year period. They can be fired if they accumulate 120 minutes of time off task in a single day or if they have accumulated 30 minutes of time off task on three separate days in a one-year period.”[25] Activities considered “time off task” include going to the bathroom, talking to another worker, or going to the wrong workstation. Workers reported that they were afraid to go to the bathroom or get a drink of water for fear of being disciplined.[26] At the end of each shift, supervisors are required to interrogate the worker with the highest time off task. In addition to handheld devices, Amazon uses an AI camera system trained on each workstation analyzing workers’ movements. The cameras automatically register the location of products and catalog every mistake workers make.[27] Monitoring the workers’ non-stop labor also helps improve the AI computer system, which learns from the responses of Amazon’s video reviewers and becomes more accurate over time.[28]
Oxfam, an international organization focused on fighting global poverty, investigated the workplace surveillance practices at Amazon and Walmart warehouses in the United States.
Employers like Amazon often claim that their surveillance systems are designed to make workers safer. “However, in recent years worker groups have decried the high injury rates and
horrific working conditions that workers encounter as Amazon employees.”[29] The report
describes the surveillance technology as follows:
The scanners play a key role in the surveillance machine because what the scanner records
can lead to “Associate Development and Performance Trackers,” or “ADAPTs,” which are
automated write-ups that penalize workers for not meeting production goals. In addition,
hundreds of security cameras are constantly monitoring the warehouse floor, ready to notify a
manager when a worker is away from their station for too long.
[. . .]
Another example of the detailed metrics that Amazon monitors is a worker’s units per hour
(UPH) score, which records how many actions a worker is able to accomplish in an hour. . . .
[W]orker metrics are prominently displayed on a monitor, which keeps workers
psychologically primed to constantly worry about “making rate” and about how they are
doing compared with their co-workers. . . . Importantly, workers are not told what the data
that electronic devices are constantly collecting is being used for, nor are they properly
notified of their privacy rights.
- What this bill would do. This bill would prohibit an “employer,” or a vendor providing services to an employer, from using the personal information of a “worker” to train an AI system to replicate, automate, or replace a worker’s job, or from providing a third party with access to a worker’s personal information to do the same. The bill defines employer to include both private- and public-sector entities, as well as an employer’s labor contractor. The bill defines “worker” to include an independent contractor providing service to or through an employer. The bill additionally requires that any contract between an employer and a vendor for services involving access to a worker’s personal information must include reasonable security procedures to protect that information. The bill’s requirements may be enforced by the Labor Commissioner or a public prosecutor; alternatively, a worker or their exclusive representative may bring a civil action for damages and other remedies.
This bill is part of a broader package of workers’ rights legislation introduced this year to address the growing impact of AI in the workplace. Other measures in this package include:
- AB 1883 (Bryan), related to workplace surveillance.
- AB 1898 (Schultz), related to employment decisions made by AI systems.
- AB 2545 (Schiavo), related to studying the impact of AI on the workforce.
- AB 2563 (Lee), related to state AI procurement standards.
- AB 2656 (Petrie-Norris), related to technological displacement of public sector workers.
- SB 947 (McNerney), related to employment decisions made by AI systems.
- SB 951 (Reyes), related to technological displacement of workers.
Together, these bills represent a coordinated effort to ensure that technological advancement does not come at the expense of workers’ dignity and safety.
4) Opposition concerns. Writing on behalf of a coalition of trade associations, the California Chamber of Commerce opposes the bill on the basis that it would “impose sweeping and unworkable restrictions on the use of worker data that will significantly hinder responsible innovation and expose employers to substantial litigation risk.” Much of the language called into question by this letter has been heavily amended or removed from the bill since the letter was submitted. Criticism that remains relevant emphasizes the following points:
CalChamber: AB 2027’s Overly Broad Restrictions on Vendor Sharing Interferes with Litigation and Compliance. Proposed Section 1571(d) (1) would prohibit vendors from transferring, disclosing, or otherwise providing access to worker data to third parties for purposes related to training AI systems. This sweeping language may unintentionally restrict routine and necessary disclosures that occur in the ordinary course of business, including disclosures required for litigation, investigations, and legal compliance . . . Because the provision would broadly restrict vendors from transmitting or otherwise making worker data available through automated or networked technologies, it would create ambiguity regarding whether vendors may continue to provide litigation-related support services that require access to workforce data.
Recent amendments narrow this bill’s focus to a worker’s “personal information,” rather than the broader “worker data,” and as a result, this concern may no longer apply. It is hard to imagine that “routine and necessary disclosure[s] that occur in the ordinary course of business” includes using a worker’s personal information to train an AI system capable of replacing them.
CalChamber: AB 2027’s Restrictions on AI Development Discourages Innovation and Workplace Improvements. AB 2027 would prohibit employers from using worker data to train AI systems intended to automate or replicate job functions. While we appreciate concerns regarding responsible deployment of AI technologies, AB 2027 fails to recognize the many beneficial uses of AI systems that assist workers rather than replace them. AI-enabled tools are frequently deployed to augment human decision-making, improve efficiency, and reduce administrative burdens that otherwise limit employee productivity.
Compared to “replacement,” which is inherently harmful to a worker being replaced, “replication” and “automation” have the potential to be genuinely beneficial – for example, the automation of a repetitive task may allow a worker to engage with more cognitively rewarding aspects of their job. Going forward, the author may wish to reconsider the strict prohibition on “automation” and “replication” that currently exists in the bill.
CalChamber: Independent Contractors Should Not Be Included in the Definition of “Worker.” The bill’s definition of “worker” includes independent contractors, which should be removed from the bill. Contractors are often limited-term workers who are performing a specific job for a company. By definition they are very different from an employee. Their contract will dictate the terms of that job and any job-specific requirements. This trend of treating them the same as employees in every new bill is at odds with prior legislation and court decisions. Further, their inclusion in such a broad bill compounds the bill’s administrative burden on businesses and state and local governments.
The vast majority of Amazon delivery drivers do not work for Amazon, despite wearing Amazon vests, driving Amazon trucks, and delivering Amazon-branded packages. They instead work for Amazon “Delivery Service Partners,” more than 4,000 independent small businesses that together represent roughly 390,000 driving jobs.[30] But while these workers do not work for Amazon, they are aggressively surveilled by the company – in 2021, each Amazon truck was equipped with an AI-powered, inward-facing camera system called Driveri, as described by the ACLU:
The device Amazon is installing (called “Driveri,” pronounced “driver eye”) has cameras pointing in four directions, one of which is toward the driver. In a video posted online, the company says the “camera records 100 percent of the time when you’re out on your route,” and watches for 16 behaviors that will “trigger Driveri to upload recorded footage.” These include not only accidents but also such things as following another car too closely, making a U-turn, failing to wear the seatbelt, obstructing the camera, “hard” braking or accelerating, and appearing to be distracted or drowsy — or what the AI interprets as those activities, anyway. Sometimes the robot camera will shout commands at you, such as “maintain safe distance!” or “please slow down!” One driver told CNBC that if the camera catches you yawning, it will tell you to pull over for at least 15 minutes — and if you don’t comply, you may get a call from your boss.[31]
Given the extensive surveillance practices employed by companies like Amazon – not only with respect to their own employees, but also independent contractors and third-party partners – it seems reasonable and appropriate for this bill to cover such circumstances.
ARGUMENTS IN SUPPORT:
The California Federations of Labor Unions, the sponsor of this measure, writes on behalf of a coalition of workers’ rights organizations:
Employers and tech companies have developed a plethora of AI tools to extract worker data. AI surveillance tools can capture all aspects of how a worker does their job relying on facial, gait, and emotional recognition and measuring neural activity. Wearable tools allow companies to analyze physical movements and not just collect data but also develop the most “efficient” movements to complete tasks. The invasiveness, scale, and velocity of AI data extraction and tool development means that as workers do their jobs, they may be unknowingly training their robot replacements. Even if they are not directly training an AI system, employers can sell worker data to brokers and tech companies for profit.
The strip mining of worker data is not simply a privacy or worker dignity issue; the scope and scale of AI surveillance and task automation threaten mass job loss across all industries. The development and deployment of AI tools to potentially replace or erode human jobs is happening at a velocity that makes it difficult for society to adapt or even mitigate the damage to workers, families, the economy, and society.
ARGUMENTS IN OPPOSITION:
California Chamber of Commerce writes in opposition to this measure on behalf of a coalition of industry trade associations:
AB 2027 would impose sweeping and unworkable restrictions on the use of worker data that
will significantly hinder responsible innovation and expose employers to substantial litigation risk. AB 2027 would undermine the ability of businesses to use common technologies that improve workplace efficiency, safety, and compliance with existing employment laws.
A coalition of organizations representing public entities, including but not limited to the California State Association of Counties and League of California Cities, writes in opposition to this measure:
We understand the reasonable concerns about the rapid integration of new technologies into daily life. However, the scope of this bill is vast and would leave significant questions about whether agencies can use AI systems or how they could train them to perform their tasks.
Local agency employers are incorporating AI and automated decision system (ADS) technology to improve service delivery, address risks, and empower employees to focus on more meaningful work through streamlining some rote aspects of job duties. State law has protected the right of public employees to bargain, and most public employees have a property interest in their position that is protected by due process. Public agencies and their workforce have long bargained over the conditions of their workplace and, more frequently, we have seen MOUs include agreements on the use of AI or ADS.
REGISTERED SUPPORT / OPPOSITION:
Support
California Labor Federation, Afl-cio (Sponsor)
Alameda Labor Council
American Federation of Musicians, Local 7
American Federation of State, County and Municipal Employees (AFSCME) California
California Alliance for Retired Americans (CARA)
California Faculty Association
California Federation of Labor Unions, Afl-cio
California Federation of Teachers Afl-cio
California School Employees Association
California State Legislative Board of the Sheet Metal, Air, Rail and Transportation Workers - Transportation Division (SMART-TD)
California Teachers Association
Central Coast Labor Council
Central Labor Council, Fresno-madera-tulare-kings Counties, Afl-cio
Inland Empire Labor Council, Afl-cio
North Bay Labor Council
North Valley Labor Federation
Oakland Privacy
Orange County Labor Federation, Afl-cio
San Mateo County Central Labor Council
Techequity Action
UAW Region 6
What We Will
Opposition
Acclamation Insurance Management Services
Allied Managed Care
Associated General Contractors of California
Associated General Contractors San Diego
Association of California Healthcare Districts (ACHD)
Association of California School Administrators
California Apartment Association
California Chamber of Commerce
California Farm Bureau
California Hospital Association
California Manufactures & Technology Association
California Retailers Association
California Special Districts Association
California State Association of Counties (CSAC)
California Trucking Association
Cinema Association of California
Civil Justice Association of California (CJAC)
Flasher Barricade Association
Leadingage California
League of California Cities
Los Angeles Area Chamber of Commerce
National Association of Professional Employer Organizations
National Payroll Reporting Consortium
Rural County Representatives of California (RCRC)
Technet
Urban Counties of California (UCC)
Oppose Unless Amended
Insights Association
Analysis Prepared by: Slater Sharp / P. & C.P. / (916) 319-2200
AB 2885 (Bauer-Kahan, Stats. 2024, Ch. 843) defined the term as “an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments.” ↑
IBM, What is machine learning?, www.ibm.com/topics/machine-learning. ↑
Ibid. ↑
OpenAI, How ChatGPT and Our Language Models Are Developed, https://help.openai.com/en/articles/7842364-how-chatgpt-and-our-foundation-models-are-developed. ↑
Netflix, How Netflix’s Recommendations System Works, https://help.netflix.com/en/node/100639. ↑
Challenger, Gray & Christmas, “Challenger Report December 2025,” (Jan. 8, 2026), https://www.challengergray.com/wp-content/uploads/2026/01/Challenger-Report-December-2025.pdf. ↑
Resume.org, “The Great Turnover: 9 in 10 Companies Plan To Hire in 2026, Yet 6 in 10 Will Have Layoffs,” https://www.resume.org/the-great-turnover-9-in-10-companies-plan-to-hire-in-2026-yet-6-in-10-will-have-layoffs-2. ↑
Salari, Beiromvand, Hosseinian-Far, Habibi, Babajani, and Mohammadi, “Impacts of generative artificial intelligence on the future of labor market: A systematic review,” Computers in Human Behavior Reports, (Mar. 15, 2025), https://www.sciencedirect.com/science/article/pii/S2451958825000673. ↑
Jeannine Mancini, “Ford CEO Says AI Will Replace 'Literally Half' of White-Collar Workers — But Blue-Collar Trades Are Still The Essential Backbone Of The Economy,” yahoo!finance, (Oct. 6, 2025), https://finance.yahoo.com/news/ford-ceo-says-ai-replace-203114506.html. ↑
Jim VandeHei and Mike Allen, “Behind the Curtain: A white-collar bloodbath,” Axios, (May 28, 2025), https://www.axios.com/2025/05/28/ai-jobs-white-collar-unemployment-anthropic. ↑
Ashton Jackson, “Nvidia CEO: You won’t lose your job to AI—you’ll ‘lose your job to somebody who uses AI’,” CNBC, (May 28, 2025), https://www.cnbc.com/2025/05/28/nvidia-ceo-jensen-huang-youll-lose-your-job-to-somebody-who-uses-ai.html. ↑
Anthropic, “Anthropic Economic Index,” https://www.anthropic.com/economic-index#state-usage. ↑
Otis College of Art and Design, Otis College Update on the Creative Economy, (Mar. 27, 2025), available at
https://www.otis.edu/about/initiatives/documents/25-063-CreativeEconomy_Report4_250325.pdf. ↑
Otis College of Art and Design, Otis College Report on the Creative Economy, (2023), available at
https://www.televisionacademy.com/files/assets/Downloads/2023-otis-report-on-the-creative-economy-v1.pdf. ↑
PMP Strategy/CISAC, Study on the economic impact of Generative AI in the Music and Audiovisual industries,
(Nov. 2024), available at https://www.cisac.org/services/reports-and-research/cisacpmp-strategy-ai-study. ↑
Katie Paul, Jeff Horwitz, and Deepa Seetharaman. “Exclusive: Meta planning sweeing layoffs as AI costs mount,” Reuters, (Mar. 13, 2026), https://www.reuters.com/business/world-at-work/meta-planning-sweeping-layoffs-ai-costs-mount-2026-03-14/. ↑
Jon Markman, “Oracle’s Massive 30,000 Layoff As AI Spending Surges,” Forbes, (Apr. 6, 2026), https://www.forbes.com/sites/jonmarkman/2026/04/06/oracles-massive-30000-layoff-as-ai-spending-surges/. ↑
Jordan Vallnsky, “Amazon is laying off 16,000 employees as AI battle intensifies,” CNN, (Jan. 28, 2026), https://www.cnn.com/2026/01/28/tech/amazon-layoffs-ai. ↑
Jack Dorsey, X, (Feb. 26, 2026), https://x.com/jack/status/2027129697092731343. ↑
Higginbotham and Wroclawski. “These Video Doorbells Have Terrible Security. Amazon Sells Them Anyway.”
Consumer Reports (Feb 29, 2024) available at https://www.consumerreports.org/home-garden/home-security-cameras/video-doorbells-sold-by-major-retailers-have-security-flaws-a2579288796/ ↑
Kevin Roose, “A Machine May Not Take Your Job, but One Could Become Your Boss,” New York Times (Jun.
23, 2019), https://www.nytimes.com/2019/06/23/technology/artificial-intelligence-ai-workplace.html ↑
Humanyze: The Future of Workforce & Market Intelligence https://humanyze.com/ ↑
Rebecca Akermann, “Your Boss is Watching,” MIT Technology Review (Feb. 24, 2025) ↑
Lauren Kaori Gurley, “Internal Documents Show Amazon’s Dystopian System for Tracking Workers Every
Minute of Their Shifts” Vice (Jun. 2, 2022), https://www.vice.com/en/article/internal-documents-show-amazons-dystopian-system-for-tracking-workers-every-minute-of-their-shifts/ ↑
Ibid. ↑
Niamh McIntyre and Rosie Bradbury, The eyes of Amazon: a hidden workforce driving a vast surveillance system,
The Bureau of Investigative Journalism (Nov. 21, 2022), https://www.thebureauinvestigates.com/stories/2022-11-21/the-eyes-of-amazon-a-hidden-workforce-driving-a-vast-surveillance-system/ ↑
Ibid. ↑
At Work and Under Watch: Surveillance and suffering at Amazon and Walmart warehouse, Oxfam (Apr. 10,
2024), https://www.oxfamamerica.org/explore/research-publications/at-work-and-under-watch/ ↑
Amazon Staff, “Everything you need to know about Amazon’s Delivery Service Partner (DSP) program,” (Oct. 28, 2024), https://www.aboutamazon.com/news/policy-news-views/amazon-dsp-program-update. ↑
Jay Stanley, “Amazon Drivers Placed Under Robot Surveillance Microscope,” (Mar. 23, 2021), https://www.aclu.org/news/privacy-technology/amazon-drivers-placed-under-robot-surveillance-microscope. ↑
Date of Hearing: April 16, 2026
Fiscal: Yes
ASSEMBLY COMMITTEE ON PRIVACY AND CONSUMER PROTECTION
Rebecca Bauer-Kahan, Chair
ABPCA Bill Id:AB 2027 (
Author:Ward) – As Amended Ver:April 9, 2026
SUBJECT: Worker data: prohibitions: artificial intelligence
SYNOPSIS
As surveillance technologies have become cheaper, more scalable, and easier to deploy, worker privacy has steadily eroded across a wide range of industries. Employers can now monitor keystrokes, employee location, communications, productivity metrics, and even biometric signals with relative ease, even extending these practices to contractors and third-party workers. Once collected, this workplace data does not go unused – it can be aggregated, labeled, and repurposed as training data for artificial intelligence (AI) systems designed to optimize workflows, automate tasks, and, in some cases, replace the very workers from whom the data was derived.
This bill would prohibit an employer, or a third-party vendor that contracts with an employer, from using a worker’s personal information to train an AI system to replicate, automate, or replace the worker’s job – either directly, or by providing a third party with access to the data. The bill would additionally require contracts between employers and third-party vendors to contain reasonable security measures to protect workers’ personal information. The bill is enforceable by the Labor Commissioner, public prosecutors, or through a civil action brought by a worker or their exclusive representative.
AB 2027 is sponsored by the California Federation of Labor Unions and supported by a variety of workers’ rights and privacy advocacy groups. It is opposed by a collection of trade associations including the California Chamber of Commerce.
EXISTING LAW:
- Provides, pursuant to the California Constitution, that all people are free and independent by nature and have inalienable rights. Among these is the fundamental right to privacy. (Cal. Const. art. I, § 1.)
- Establishes the Information Practices Act of 1977, governing the collection, maintenance, use, and dissemination of personal information by California state agencies. (Civ. Code § 1798 et seq.)
- Provides the following definition for “personal information”:
(1) “Personal information” means information that identifies, relates to, describes, is reasonably capable of being associated with, or could reasonably be linked, directly or indirectly, with a particular consumer or household. Personal information includes, but is not limited to, the following if it identifies, relates to, describes, is reasonably capable of being associated with, or could be reasonably linked, directly or indirectly, with a particular consumer or household:
(A) Identifiers such as a real name, alias, postal address, unique personal identifier, online identifier, Internet Protocol address, email address, account name, social security number, driver’s license number, passport number, or other similar identifiers.
(B) Any personal information described in subdivision (e) of Section 1798.80.
(C) Characteristics of protected classifications under California or federal law.
(D) Commercial information, including records of personal property, products or services purchased, obtained, or considered, or other purchasing or consuming histories or tendencies.
(E) Biometric information.
(F) Internet or other electronic network activity information, including, but not limited to, browsing history, search history, and information regarding a consumer’s interaction with an internet website application, or advertisement.
(G) Geolocation data.
(H) Audio, electronic, visual, thermal, olfactory, or similar information.
(I) Professional or employment-related information.
(J) Education information, defined as information that is not publicly available personally identifiable information as defined in the Family Educational Rights and Privacy Act (20 U.S.C. Sec. 1232g; 34 C.F.R. Part 99).
(K) Inferences drawn from any of the information identified in this subdivision to create a profile about a consumer reflecting the consumer’s preferences, characteristics, psychological trends, predispositions, behavior, attitudes, intelligence, abilities, and aptitudes.
(L) Sensitive personal information.
(2) (A) “Personal information” does not include publicly available information or lawfully obtained, truthful information that is a matter of public concern.
(B) (i) For purposes of this paragraph, “publicly available” means any of the following:
(I) Information that is lawfully made available from federal, state, or local government records.
(II) Information that a business has a reasonable basis to believe is lawfully made available to the general public by the consumer or from widely distributed media.
(III) Information made available by a person to whom the consumer has disclosed the information if the consumer has not restricted the information to a specific audience.
(ii) “Publicly available” does not mean biometric information collected by a business about a consumer without the consumer’s knowledge.
(3) “Personal information” does not include consumer information that is deidentified or aggregate consumer information.
(4) “Personal information” can exist in various formats, including, but not limited to, all of the following:
(A) Physical formats, including paper documents, printed images, vinyl records, or video tapes.
(B) Digital formats, including text, image, audio, or video files.
(C) Abstract digital formats, including compressed or encrypted files, metadata, or artificial intelligence systems that are capable of outputting personal information. (Civ. Code § 1798.140.)
- Establishes the California Consumer Privacy Act, which grants consumers certain rights regarding their personal information, including enhanced notice, access, and disclosure; the right to deletion; the right to restrict the sale of information; and protection from discrimination for exercising these rights. (Civ. Code § 1798.100 et seq.)
- Establishes the Division of Labor Standards Enforcement (DLSE), under the direction of the Labor Commissioner, within the Department of Industrial Relations, and authorizes the LC to investigate employee complaints and enforce labor laws, as specified. (Lab. Code § 79 et seq.)
- Authorizes the DLSE to enforce the Labor Code and all labor laws of the state, the enforcement of which is not specifically vested in any other officer, board, or commission. (Lab. Code § 95(a).)
- Authorizes the Labor Commissioner to prosecute actions for the collection of wages, penalties, and demands, as specified. (Lab. Code § 98.3.)
- Authorizes, until January 1, 2029, the Attorney General, a district attorney, a city attorney, a county counsel, or any other city or county prosecutor, to prosecute an action, either civil or criminal, for violation of certain provisions of the Labor Code or to enforce those provisions independently. (Lab. Code § 180 et seq.)
THIS BILL:
- Defines:
- “Artificial intelligence” to mean an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments.
- “Employer” to mean a person, business, private entity, state, city, county, charter county, city and county, municipality, charter municipality, special district, transit district, the University of California, the California State University, community college district, school district, or any other state or local governmental entity that, directly or indirectly, or through an agent or any other person, employs or exercises control over the wages, benefits, other compensation, hours, working conditions, access to work or job opportunities, or other terms or conditions of employment, of any worker. Further defines “employer” to include an employer’s labor contractor.
- “Personal information” to have the same meaning as defined in Section 1798.140 of the Civil Code.
- “Public prosecutor” to have the same meaning as defined in Section 180 of the Labor Code.
- “Worker” to mean a natural person who is an employee of, or an independent contractor providing service to or through, an employer.
- Prohibits an employer from using a worker’s personal information to train an artificial intelligence system to replicate, automate, or replace a worker’s job.
- Prohibits an employer from selling, disclosing, licensing, transmitting, making available, exposing, processing, monetizing, or otherwise providing access to a worker’s personal information to a third party for the purpose of training an artificial intelligence system to replicate, automate, or replace a worker’s job.
- Prohibits a vendor from providing services to an employer under a contract from doing any of the following:
- Transferring, selling, disclosing, licensing, transmitting, making available, exposing, processing, monetizing, or otherwise providing access to the personal information of an employer’s worker to a third party for the purpose of training an artificial intelligence system to replicate, automate, or replace a worker’s job.
- Using the personal information of an employer’s worker to train an artificial intelligence system to replicate, automate, or replace a worker’s job if the vendor sells or leases the artificial intelligence system to other customers.
- Requires that a contract between an employer and a vendor for services involving the access or capture of a worker’s personal information on the employer’s behalf include a requirement that the vendor must implement and maintain reasonable security procedures to protect the worker’s personal information from unauthorized or illegal access, destruction, use, modification, or disclosure.
- Provides that the Labor Commissioner or a public prosecutor shall enforce the provisions of the bill, or that a worker or their exclusive representative who suffers a violation may bring a civil action, and authorizes a penalty of up to $500 for each violation in addition to any injunctive relief, punitive damages, and reasonable attorney’s fees and costs authorized by the court. Specifies that actions may be brought in the superior court in any county in which the violation in question is alleged to have occurred, or in which the employer resides. Specifies that enforcement provisions do not preempt any city, county, or city and county ordinance that provides equal or greater protection to workers.
COMMENTS:
- Author’s statement. According to the author:
Artificial intelligence is rapidly reshaping our economy, but it is increasingly being built on the backs of workers without their knowledge or consent. The data workers generate every day is fueling AI systems that can monitor, analyze, and ultimately replace them, while employers and tech companies profit from its extraction and sale.
While workers have historically faced being asked to train their replacements, the scale and speed of today’s technology is unprecedented. AI systems can now learn from human labor in real time, accelerating job loss and deskilling across industries, from call centers to software engineering. This is not just about privacy or dignity, it is a direct threat to jobs and economic stability.
AB 2027 sets clear boundaries. It prohibits employers from using worker data to train AI systems that automate or replace jobs, and prevents the sale or sharing of that data for that purpose. It also ensures tech developers cannot use worker data collected through their tools to build systems that displace workers.
This bill affirms a simple principle: workers should not be used to train their own replacements, and innovation must not come at the expense of economic security.
- Background. Artificial intelligence. AI refers to the mimicking of human intelligence by artificial systems such as computers.[1] AI uses algorithms – sets of rules – to transform inputs into outputs. Inputs and outputs can be anything a computer is able to process: numbers, text, audio, or video, for example. AI is not fundamentally different from other computer functions; unlike other computer functions, however, AI is able to accomplish tasks that are normally performed by humans.
Most modern AI tools are created through a process known as “machine learning,” which enables AI tools to learn relationships between inputs and outputs without being explicitly programmed.[2] The process of exposing a naïve AI to data is known as “training.” The algorithm that an AI develops during training is known as its “model.” At its core, training is an optimization problem: machine learning attempts to identify model parameters – weights – that minimize the difference between predicted outcomes and actual outcomes. During training, these weights are continuously adjusted to improve the model’s performance. Once trained, the model can process new, never-before-seen data.[3]
Models trained on small, specific datasets in order to make recommendations and predictions are sometimes referred to as “predictive AI.” This differentiates them from generative AI (GenAI) which are trained on massive datasets in order to produce detailed text, images, audio, and video. When ChatGPT generates text in clear, concise paragraphs, it uses GenAI that has been trained on the written contents of the internet.[4] When Netflix suggests content to a viewer, its recommendation is produced using predictive AI that has been trained on the viewing habits of Netflix users.[5]
AI’s effect on labor markets. According to a recent report from executive outplacement firm Challenger, Gray & Christmas, AI was responsible for 54,836 announced layoffs in the United States in 2025.[6] Earlier this year, a Resume.org survey of 1000 United States hiring managers revealed that 55% of companies expected further layoffs in 2026, and that 44% of these companies cited AI as a top driver – though 59% admitted they intentionally emphasized AI when explaining hiring freezes or layoffs because it “plays better with stakeholders.” [7] It is difficult to separate the signal from the noise when assessing the effect that advanced AI is having on the labor market, as emphasized in a recent review published in Computers in Human Behavior Reports:
The use of Generative AI has significant effects on businesses and the labor market. One of the key contributions of this technology is the ability to generate content. However, the generation of inaccurate and incomplete content by AI may lead to a decrease in trust in the technology. In some industries and occupations, Generative AI may replace employees in content creation activities and related tasks, reducing job opportunities in these sectors. On the other hand, GenAI may create new specializations in content creation, leading to new job opportunities for people with AI-related knowledge.[8]
Ford CEO Jim Farley recently expressed his belief that “artificial intelligence is going to replace literally half of all white collar workers,”[9] a belief shared by Anthropic CEO Darius Amodei.[10] On the other hand, Nvidia CEO Jensen Huang recently predicted that AI will primarily augment, rather than replace, American workers.[11] Anthropic’s “Economic Index” tool, which tracks AI usage trends across the workforce, reveals that California currently ranks fifth among states in terms of AI usage, behind Washington DC, Massachusetts, Washington state, and New York.[12]
California’s creative economy. The creative economy is both hugely important to California and particularly sensitive to AI. The Otis College 2025 Report on the Creative Economy estimates that California’s creative economy employs more than 760,000 workers with an average salary of $191,000 per year, accounting for roughly 5% of the state’s workforce.[13] The 2023 version of this report estimated that California’s creative economy was responsible for 14.9% of the state’s $3.4 trillion economy in 2021, generating over $194.1 billion in federal, state, and municipal taxes.[14]
While the adoption of GenAI is opening up new possibilities in music, film, publishing and other creative fields, the technology’s impact on California’s creative economy and individual creators remains uncertain and unevenly distributed. To date, few studies have assessed how AI will affect jobs and incomes for writers, actors, musicians, designers and other creators. A global report published by the International Confederation of Societies of Authors and Composers (CISAC and PMP Strategy) in December of 2024 projected the emergence of a massive market for AI-generated content. The study estimated that AI-generated audiovisual content, including film and TV, could be worth €48 billion ($55.73 billion) by 2028. However, the report warned that up to 21% of creators’ revenue in film and TV could be cannibalized as a result:
In an unchanged regulatory framework, creators will actually suffer losses on two fronts: the loss of revenues due to the unauthorised use of their works by Gen AI models without remuneration; and replacement of their traditional revenue streams due to the substitution effect of AI-generated outputs, competing against human- made works.[15]
Tech industry layoffs. Over the past two years, major California-based tech companies have laid off hundreds of thousands of workers around the globe. These companies have increasingly linked these layoffs to AI: in March, Reuters reported that Meta’s layoff plans, which could affect 20% or more of the company, were intended to offset AI infrastructure investments and prepare for greater efficiency brought about by AI-assisted workers.[16] Oracle similarly laid off 30,000 employees in late March – representing roughly 18% of its global workforce – in order to fund its commitment to aggressively build out AI infrastructure.[17] Amazon announced plans to lay off 16,000 employees in January of this year, its second round of large-scale job reductions in three months. The company announced in a blog post that the layoffs were necessary to increase its decision-making speed.[18] Jack Dorsey, founder and CEO of startup “Block, Inc” tied cutting 40% of his company to advances in efficiency due to AI in a recent X post:
we're not making this decision because we're in trouble. our business is strong. gross profit continues to grow, we continue to serve more and more customers, and profitability is improving. but something has changed. we're already seeing that the intelligence tools we’re creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company. and that's accelerating rapidly.[19]
Workplace surveillance. Over the last five years, surveillance tools have become more affordable and more intrusive. Employers can collect vast dossiers on their employees, gathering sweeping amounts of data about every aspect of their jobs and personal lives. Often this is done “without employees’ full informed or free consent. Many workers, while generally aware they are being monitored, don’t know the extent of the surveillance or what is being done with the information.”[20]
Employers are using more surveillance technology than ever — digital cameras, motion scanners, RFID badges, Apple Watch badges, Bluetooth beacons, keystroke logging — to track
every single movement of workers in the office and to gauge their productivity. Some workplaces are using biometric data such as eye movements, body shifts, and facial expressions,
captured by computer webcams, to evaluate whether their employees are being appropriately
attentive in their work tasks. As an example, AI systems at call centers record and grade how workers are handling calls. This technology can be used to “coach” workers while they are talking to customers, telling them to sound happier or be more empathetic.[21] Another example is wearable technology that, among other things, tracks a worker’s movements throughout the day, gathering biometric data, measuring how many times they use the bathroom, how long they spend in break areas, and which employees are spending time together. At least one company sells biometric ID badges with microphones, sensors, and other tools to record conversations, monitor speech, body movements, and location.[22] Even body temperature, sweat, and frequency of bathroom visits can be tracked and analyzed by employers.
A recent article in MIT Technology Review describes one company’s surveillance tool this way:
Companies that use electronic employee monitoring report that they are most often looking to the technologies not only to increase productivity but also to manage risk. And software like Teramind[23] offers tools and analysis to help with both priorities. While Teramind, a globally distributed company, keeps its list of over 10,000 client companies private, it provides resources for the financial, health-care, and customer service industries, among others—some of which have strict compliance requirements that can be tricky to keep on top of. The platform allows clients to set data-driven standards for productivity, establish thresholds for alerts about toxic communication tone or language, create tracking systems for sensitive file sharing, and more.
[. . .]
Selecting and tuning the appropriate combination of data is up to Teramind’s clients and depends on the size, goals, and capabilities of the particular company. The companies are also the ones to decide, based on their legal and compliance requirements, what measures to take if thresholds for negative behavior or low performance are hit.[24]
According to documents filed by Amazon workers with the National Labor Relations Board, Amazon tracks every minute that their workers spend not performing their assigned tasks. To facilitate this, Amazon uses handheld scanners that are also used to track packages. One worker claims they “can receive a written warning for accumulating 30 minutes of time off task in a day one time in a rolling one-year period. They can be fired if they accumulate 120 minutes of time off task in a single day or if they have accumulated 30 minutes of time off task on three separate days in a one-year period.”[25] Activities considered “time off task” include going to the bathroom, talking to another worker, or going to the wrong workstation. Workers reported that they were afraid to go to the bathroom or get a drink of water for fear of being disciplined.[26] At the end of each shift, supervisors are required to interrogate the worker with the highest time off task. In addition to handheld devices, Amazon uses an AI camera system trained on each workstation analyzing workers’ movements. The cameras automatically register the location of products and catalog every mistake workers make.[27] Monitoring the workers’ non-stop labor also helps improve the AI computer system, which learns from the responses of Amazon’s video reviewers and becomes more accurate over time.[28]
Oxfam, an international organization focused on fighting global poverty, investigated the workplace surveillance practices at Amazon and Walmart warehouses in the United States.
Employers like Amazon often claim that their surveillance systems are designed to make workers safer. “However, in recent years worker groups have decried the high injury rates and
horrific working conditions that workers encounter as Amazon employees.”[29] The report
describes the surveillance technology as follows:
The scanners play a key role in the surveillance machine because what the scanner records
can lead to “Associate Development and Performance Trackers,” or “ADAPTs,” which are
automated write-ups that penalize workers for not meeting production goals. In addition,
hundreds of security cameras are constantly monitoring the warehouse floor, ready to notify a
manager when a worker is away from their station for too long.
[. . .]
Another example of the detailed metrics that Amazon monitors is a worker’s units per hour
(UPH) score, which records how many actions a worker is able to accomplish in an hour. . . .
[W]orker metrics are prominently displayed on a monitor, which keeps workers
psychologically primed to constantly worry about “making rate” and about how they are
doing compared with their co-workers. . . . Importantly, workers are not told what the data
that electronic devices are constantly collecting is being used for, nor are they properly
notified of their privacy rights.
- What this bill would do. This bill would prohibit an “employer,” or a vendor providing services to an employer, from using the personal information of a “worker” to train an AI system to replicate, automate, or replace a worker’s job, or from providing a third party with access to a worker’s personal information to do the same. The bill defines employer to include both private- and public-sector entities, as well as an employer’s labor contractor. The bill defines “worker” to include an independent contractor providing service to or through an employer. The bill additionally requires that any contract between an employer and a vendor for services involving access to a worker’s personal information must include reasonable security procedures to protect that information. The bill’s requirements may be enforced by the Labor Commissioner or a public prosecutor; alternatively, a worker or their exclusive representative may bring a civil action for damages and other remedies.
This bill is part of a broader package of workers’ rights legislation introduced this year to address the growing impact of AI in the workplace. Other measures in this package include:
- AB 1883 (Bryan), related to workplace surveillance.
- AB 1898 (Schultz), related to employment decisions made by AI systems.
- AB 2545 (Schiavo), related to studying the impact of AI on the workforce.
- AB 2563 (Lee), related to state AI procurement standards.
- AB 2656 (Petrie-Norris), related to technological displacement of public sector workers.
- SB 947 (McNerney), related to employment decisions made by AI systems.
- SB 951 (Reyes), related to technological displacement of workers.
Together, these bills represent a coordinated effort to ensure that technological advancement does not come at the expense of workers’ dignity and safety.
4) Opposition concerns. Writing on behalf of a coalition of trade associations, the California Chamber of Commerce opposes the bill on the basis that it would “impose sweeping and unworkable restrictions on the use of worker data that will significantly hinder responsible innovation and expose employers to substantial litigation risk.” Much of the language called into question by this letter has been heavily amended or removed from the bill since the letter was submitted. Criticism that remains relevant emphasizes the following points:
CalChamber: AB 2027’s Overly Broad Restrictions on Vendor Sharing Interferes with Litigation and Compliance. Proposed Section 1571(d) (1) would prohibit vendors from transferring, disclosing, or otherwise providing access to worker data to third parties for purposes related to training AI systems. This sweeping language may unintentionally restrict routine and necessary disclosures that occur in the ordinary course of business, including disclosures required for litigation, investigations, and legal compliance . . . Because the provision would broadly restrict vendors from transmitting or otherwise making worker data available through automated or networked technologies, it would create ambiguity regarding whether vendors may continue to provide litigation-related support services that require access to workforce data.
Recent amendments narrow this bill’s focus to a worker’s “personal information,” rather than the broader “worker data,” and as a result, this concern may no longer apply. It is hard to imagine that “routine and necessary disclosure[s] that occur in the ordinary course of business” includes using a worker’s personal information to train an AI system capable of replacing them.
CalChamber: AB 2027’s Restrictions on AI Development Discourages Innovation and Workplace Improvements. AB 2027 would prohibit employers from using worker data to train AI systems intended to automate or replicate job functions. While we appreciate concerns regarding responsible deployment of AI technologies, AB 2027 fails to recognize the many beneficial uses of AI systems that assist workers rather than replace them. AI-enabled tools are frequently deployed to augment human decision-making, improve efficiency, and reduce administrative burdens that otherwise limit employee productivity.
Compared to “replacement,” which is inherently harmful to a worker being replaced, “replication” and “automation” have the potential to be genuinely beneficial – for example, the automation of a repetitive task may allow a worker to engage with more cognitively rewarding aspects of their job. Going forward, the author may wish to reconsider the strict prohibition on “automation” and “replication” that currently exists in the bill.
CalChamber: Independent Contractors Should Not Be Included in the Definition of “Worker.” The bill’s definition of “worker” includes independent contractors, which should be removed from the bill. Contractors are often limited-term workers who are performing a specific job for a company. By definition they are very different from an employee. Their contract will dictate the terms of that job and any job-specific requirements. This trend of treating them the same as employees in every new bill is at odds with prior legislation and court decisions. Further, their inclusion in such a broad bill compounds the bill’s administrative burden on businesses and state and local governments.
The vast majority of Amazon delivery drivers do not work for Amazon, despite wearing Amazon vests, driving Amazon trucks, and delivering Amazon-branded packages. They instead work for Amazon “Delivery Service Partners,” more than 4,000 independent small businesses that together represent roughly 390,000 driving jobs.[30] But while these workers do not work for Amazon, they are aggressively surveilled by the company – in 2021, each Amazon truck was equipped with an AI-powered, inward-facing camera system called Driveri, as described by the ACLU:
The device Amazon is installing (called “Driveri,” pronounced “driver eye”) has cameras pointing in four directions, one of which is toward the driver. In a video posted online, the company says the “camera records 100 percent of the time when you’re out on your route,” and watches for 16 behaviors that will “trigger Driveri to upload recorded footage.” These include not only accidents but also such things as following another car too closely, making a U-turn, failing to wear the seatbelt, obstructing the camera, “hard” braking or accelerating, and appearing to be distracted or drowsy — or what the AI interprets as those activities, anyway. Sometimes the robot camera will shout commands at you, such as “maintain safe distance!” or “please slow down!” One driver told CNBC that if the camera catches you yawning, it will tell you to pull over for at least 15 minutes — and if you don’t comply, you may get a call from your boss.[31]
Given the extensive surveillance practices employed by companies like Amazon – not only with respect to their own employees, but also independent contractors and third-party partners – it seems reasonable and appropriate for this bill to cover such circumstances.
ARGUMENTS IN SUPPORT:
The California Federations of Labor Unions, the sponsor of this measure, writes on behalf of a coalition of workers’ rights organizations:
Employers and tech companies have developed a plethora of AI tools to extract worker data. AI surveillance tools can capture all aspects of how a worker does their job relying on facial, gait, and emotional recognition and measuring neural activity. Wearable tools allow companies to analyze physical movements and not just collect data but also develop the most “efficient” movements to complete tasks. The invasiveness, scale, and velocity of AI data extraction and tool development means that as workers do their jobs, they may be unknowingly training their robot replacements. Even if they are not directly training an AI system, employers can sell worker data to brokers and tech companies for profit.
The strip mining of worker data is not simply a privacy or worker dignity issue; the scope and scale of AI surveillance and task automation threaten mass job loss across all industries. The development and deployment of AI tools to potentially replace or erode human jobs is happening at a velocity that makes it difficult for society to adapt or even mitigate the damage to workers, families, the economy, and society.
ARGUMENTS IN OPPOSITION:
California Chamber of Commerce writes in opposition to this measure on behalf of a coalition of industry trade associations:
AB 2027 would impose sweeping and unworkable restrictions on the use of worker data that
will significantly hinder responsible innovation and expose employers to substantial litigation risk. AB 2027 would undermine the ability of businesses to use common technologies that improve workplace efficiency, safety, and compliance with existing employment laws.
A coalition of organizations representing public entities, including but not limited to the California State Association of Counties and League of California Cities, writes in opposition to this measure:
We understand the reasonable concerns about the rapid integration of new technologies into daily life. However, the scope of this bill is vast and would leave significant questions about whether agencies can use AI systems or how they could train them to perform their tasks.
Local agency employers are incorporating AI and automated decision system (ADS) technology to improve service delivery, address risks, and empower employees to focus on more meaningful work through streamlining some rote aspects of job duties. State law has protected the right of public employees to bargain, and most public employees have a property interest in their position that is protected by due process. Public agencies and their workforce have long bargained over the conditions of their workplace and, more frequently, we have seen MOUs include agreements on the use of AI or ADS.
REGISTERED SUPPORT / OPPOSITION:
Support
California Labor Federation, Afl-cio (Sponsor)
Alameda Labor Council
American Federation of Musicians, Local 7
American Federation of State, County and Municipal Employees (AFSCME) California
California Alliance for Retired Americans (CARA)
California Faculty Association
California Federation of Labor Unions, Afl-cio
California Federation of Teachers Afl-cio
California School Employees Association
California State Legislative Board of the Sheet Metal, Air, Rail and Transportation Workers - Transportation Division (SMART-TD)
California Teachers Association
Central Coast Labor Council
Central Labor Council, Fresno-madera-tulare-kings Counties, Afl-cio
Inland Empire Labor Council, Afl-cio
North Bay Labor Council
North Valley Labor Federation
Oakland Privacy
Orange County Labor Federation, Afl-cio
San Mateo County Central Labor Council
Techequity Action
UAW Region 6
What We Will
Opposition
Acclamation Insurance Management Services
Allied Managed Care
Associated General Contractors of California
Associated General Contractors San Diego
Association of California Healthcare Districts (ACHD)
Association of California School Administrators
California Apartment Association
California Chamber of Commerce
California Farm Bureau
California Hospital Association
California Manufactures & Technology Association
California Retailers Association
California Special Districts Association
California State Association of Counties (CSAC)
California Trucking Association
Cinema Association of California
Civil Justice Association of California (CJAC)
Flasher Barricade Association
Leadingage California
League of California Cities
Los Angeles Area Chamber of Commerce
National Association of Professional Employer Organizations
National Payroll Reporting Consortium
Rural County Representatives of California (RCRC)
Technet
Urban Counties of California (UCC)
Oppose Unless Amended
Insights Association
Analysis Prepared by: Slater Sharp / P. & C.P. / (916) 319-2200
AB 2885 (Bauer-Kahan, Stats. 2024, Ch. 843) defined the term as “an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments.” ↑
IBM, What is machine learning?, www.ibm.com/topics/machine-learning. ↑
Ibid. ↑
OpenAI, How ChatGPT and Our Language Models Are Developed, https://help.openai.com/en/articles/7842364-how-chatgpt-and-our-foundation-models-are-developed. ↑
Netflix, How Netflix’s Recommendations System Works, https://help.netflix.com/en/node/100639. ↑
Challenger, Gray & Christmas, “Challenger Report December 2025,” (Jan. 8, 2026), https://www.challengergray.com/wp-content/uploads/2026/01/Challenger-Report-December-2025.pdf. ↑
Resume.org, “The Great Turnover: 9 in 10 Companies Plan To Hire in 2026, Yet 6 in 10 Will Have Layoffs,” https://www.resume.org/the-great-turnover-9-in-10-companies-plan-to-hire-in-2026-yet-6-in-10-will-have-layoffs-2. ↑
Salari, Beiromvand, Hosseinian-Far, Habibi, Babajani, and Mohammadi, “Impacts of generative artificial intelligence on the future of labor market: A systematic review,” Computers in Human Behavior Reports, (Mar. 15, 2025), https://www.sciencedirect.com/science/article/pii/S2451958825000673. ↑
Jeannine Mancini, “Ford CEO Says AI Will Replace 'Literally Half' of White-Collar Workers — But Blue-Collar Trades Are Still The Essential Backbone Of The Economy,” yahoo!finance, (Oct. 6, 2025), https://finance.yahoo.com/news/ford-ceo-says-ai-replace-203114506.html. ↑
Jim VandeHei and Mike Allen, “Behind the Curtain: A white-collar bloodbath,” Axios, (May 28, 2025), https://www.axios.com/2025/05/28/ai-jobs-white-collar-unemployment-anthropic. ↑
Ashton Jackson, “Nvidia CEO: You won’t lose your job to AI—you’ll ‘lose your job to somebody who uses AI’,” CNBC, (May 28, 2025), https://www.cnbc.com/2025/05/28/nvidia-ceo-jensen-huang-youll-lose-your-job-to-somebody-who-uses-ai.html. ↑
Anthropic, “Anthropic Economic Index,” https://www.anthropic.com/economic-index#state-usage. ↑
Otis College of Art and Design, Otis College Update on the Creative Economy, (Mar. 27, 2025), available at
https://www.otis.edu/about/initiatives/documents/25-063-CreativeEconomy_Report4_250325.pdf. ↑
Otis College of Art and Design, Otis College Report on the Creative Economy, (2023), available at
https://www.televisionacademy.com/files/assets/Downloads/2023-otis-report-on-the-creative-economy-v1.pdf. ↑
PMP Strategy/CISAC, Study on the economic impact of Generative AI in the Music and Audiovisual industries,
(Nov. 2024), available at https://www.cisac.org/services/reports-and-research/cisacpmp-strategy-ai-study. ↑
Katie Paul, Jeff Horwitz, and Deepa Seetharaman. “Exclusive: Meta planning sweeing layoffs as AI costs mount,” Reuters, (Mar. 13, 2026), https://www.reuters.com/business/world-at-work/meta-planning-sweeping-layoffs-ai-costs-mount-2026-03-14/. ↑
Jon Markman, “Oracle’s Massive 30,000 Layoff As AI Spending Surges,” Forbes, (Apr. 6, 2026), https://www.forbes.com/sites/jonmarkman/2026/04/06/oracles-massive-30000-layoff-as-ai-spending-surges/. ↑
Jordan Vallnsky, “Amazon is laying off 16,000 employees as AI battle intensifies,” CNN, (Jan. 28, 2026), https://www.cnn.com/2026/01/28/tech/amazon-layoffs-ai. ↑
Jack Dorsey, X, (Feb. 26, 2026), https://x.com/jack/status/2027129697092731343. ↑
Higginbotham and Wroclawski. “These Video Doorbells Have Terrible Security. Amazon Sells Them Anyway.”
Consumer Reports (Feb 29, 2024) available at https://www.consumerreports.org/home-garden/home-security-cameras/video-doorbells-sold-by-major-retailers-have-security-flaws-a2579288796/ ↑
Kevin Roose, “A Machine May Not Take Your Job, but One Could Become Your Boss,” New York Times (Jun.
23, 2019), https://www.nytimes.com/2019/06/23/technology/artificial-intelligence-ai-workplace.html ↑
Humanyze: The Future of Workforce & Market Intelligence https://humanyze.com/ ↑
Rebecca Akermann, “Your Boss is Watching,” MIT Technology Review (Feb. 24, 2025) ↑
Lauren Kaori Gurley, “Internal Documents Show Amazon’s Dystopian System for Tracking Workers Every
Minute of Their Shifts” Vice (Jun. 2, 2022), https://www.vice.com/en/article/internal-documents-show-amazons-dystopian-system-for-tracking-workers-every-minute-of-their-shifts/ ↑
Ibid. ↑
Niamh McIntyre and Rosie Bradbury, The eyes of Amazon: a hidden workforce driving a vast surveillance system,
The Bureau of Investigative Journalism (Nov. 21, 2022), https://www.thebureauinvestigates.com/stories/2022-11-21/the-eyes-of-amazon-a-hidden-workforce-driving-a-vast-surveillance-system/ ↑
Ibid. ↑
At Work and Under Watch: Surveillance and suffering at Amazon and Walmart warehouse, Oxfam (Apr. 10,
2024), https://www.oxfamamerica.org/explore/research-publications/at-work-and-under-watch/ ↑
Amazon Staff, “Everything you need to know about Amazon’s Delivery Service Partner (DSP) program,” (Oct. 28, 2024), https://www.aboutamazon.com/news/policy-news-views/amazon-dsp-program-update. ↑
Jay Stanley, “Amazon Drivers Placed Under Robot Surveillance Microscope,” (Mar. 23, 2021), https://www.aclu.org/news/privacy-technology/amazon-drivers-placed-under-robot-surveillance-microscope. ↑
Date of Hearing: April 8, 2026
ASSEMBLY COMMITTEE ON LABOR AND EMPLOYMENT
Liz Ortega, Chair
ABPCA Bill Id:AB 2027 (
Author:Ward) – As Amended Ver:March 16, 2026
SUBJECT: Worker data: prohibitions: artificial intelligence
SUMMARY: Regulates the use of worker data by an employer, vendor acting on behalf of an employer, or a vendor providing services to an employer under a contract. Specifically, this bill:
- Requires an employer, or a vendor acting on behalf of an employer, to collect and process worker data only as strictly necessary to administer the employment relationship and to fulfill specific, affirmatively requested employment-related or legal obligations; and prohibits an employer or vendor acting on behalf of an employer from collecting or using worker data for any other purpose, including, but not limited to, using worker data to train an artificial intelligence (AI) system to replicate, automate, or replace a worker’s job.
- Prohibits an employer, or vendor acting on behalf of an employer, from deploying AI trained with worker data to replicate, automate, or replace a worker’s job.
- Prohibits an employer from selling, disclosing, licensing, transmitting, making available, exposing, processing, monetizing, or otherwise providing access to worker data to a third party, including through automated means, embedded technologies, tracking technologies, networked data collection, or transmission mechanisms, to train an AI system for the purposes of replicating, automating, or replacing a worker’s job.
- Prohibits a vendor providing services to an employer under a contract from doing any of the following:
- Transferring, selling, disclosing, licensing, transmitting, making available, exposing, processing, monetizing, or otherwise providing access to the employer’s worker data to a third party, including through automated means, embedded technologies, tracking technologies, or networked data collection or transmission mechanisms to train an AI system for the purposes of replicating, automating, or replacing a worker’s job.
- Using the employer’s worker data to train any of the vendor’s AI products, sold or leased to other customers, for the purpose of replicating, automating, or replacing a worker’s job.
- Requires a contract between an employer and a vendor for services involving the access or capture of worker data on the employer’s behalf to include both of the following terms:
- A requirement that the vendor must implement and maintain reasonable security procedures to protect the worker data from unauthorized or illegal access, destruction, use, modification, or disclosure.
- A provision that the vendor and employer agree to be jointly and severally liable for a breach of the vendor’s data security systems or unauthorized access, to the extent the breach or access involves worker data provided by the employer, and for any violation of this bill’s provisions.
- Requires the Labor Commissioner (LC), and authorizes a public prosecutor, to enforce the provisions of the bill, as specified.
- Authorizes, alternatively to enforcement by the LC or a public prosecutor, a worker, or their exclusive representative, who has suffered a violation to bring a civil action in a court of competent jurisdiction for damages caused by that adverse action, including punitive damages.
- Authorizes, in any civil action brought pursuant to (6)-(7) above, the petitioner to seek appropriate temporary or preliminary injunctive relief, punitive damages, and reasonable attorney’s fees and costs.
- Provides that an employer in violation may be subject to a penalty of up to $500 per worker for each violation.
- Authorizes a worker, the LC, or a public prosecutor to recover a penalty as a statutory penalty paid to the worker or a civil penalty, but not both, for the same violation.
- Authorizes an action pursuant to (6)-(10) to be brought in the superior court in any county in which the violation in question is alleged to have occurred or in which the employer resides or transacts business.
- Provides that the bill’s provisions do not preempt any city, county, or city and county ordinance that provides equal or greater protection to workers who are covered by this part.
- Applies the bill to all cities, including charter cities.
- Defines certain terms, including:
- “Artificial intelligence” to mean an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments.
- “Employer” to mean a person, business, private entity, state, city, county, charter county, city and county, municipality, charter municipality, special district, transit district, the University of California, the California State University, community college district, school district, or any other state or local governmental entity that, directly or indirectly, or through an agent or any other person, employs or exercises control over the wages, benefits, other compensation, hours, working conditions, access to work or job opportunities, or other terms or conditions of employment, of any worker. “Employer” includes an employer’s labor contractor.
- “Public prosecutor” to have the same meaning as defined in Labor Code Section 180: the Attorney General, a district attorney, a city attorney, a county counsel, or any other city or county prosecutor.
- “Worker” to mean a natural person who is an employee of, or an independent contractor providing service to or through, an employer.
- “Worker data” to mean any of the following:
- Information that identifies, relates to, describes, is reasonably capable of being associated with, or could reasonably be linked, directly or indirectly, with the personal information relating to, a worker, regardless of how the information is collected, inferred, or obtained, including, but not limited to, a worker’s biometric data, employment history, or personal identifying information.
- Information that in any way relates to or describes, regardless of how the information is collected, inferred, or obtained, how a worker performs their job duties.
EXISTING LAW:
- Establishes the Division of Labor Standards Enforcement (DLSE), under the direction of the LC, within the Department of Industrial Relations, and authorizes the LC to investigate employee complaints and enforce labor laws, as specified. Labor Code § 79 et seq.
- Authorizes the DLSE to enforce the Labor Code and all labor laws of the state, the enforcement of which is not specifically vested in any other officer, board, or commission. Labor Code § 95(a).
- Authorizes the LC to prosecute actions for the collection of wages, penalties, and demands, as specified. Labor Code § 98.3.
- Authorizes, until January 1, 2029, a public prosecutor, as defined, to prosecute an action, either civil or criminal, for violation of certain provisions of the Labor Code or to enforce those provisions independently. Labor Code § 180 et seq.
- Establishes the California Consumer Privacy Act (CCPA), which grants consumers certain rights regarding their personal information, including enhanced notice, access, and disclosure; the right to deletion; the right to restrict the sale of information; and protection from discrimination for exercising these rights. Civil Code § 1798.100 et seq.
- Establishes the Consumer Privacy Rights Act (CPRA), which amends the CCPA and creates the California Privacy Protection Agency, which is charged with implementing these privacy laws, promulgating regulations, and carrying out enforcement actions. Civil Code § 1798.100 et seq.; Proposition 24 (2020).
FISCAL EFFECT: Unknown
COMMENTS: Note: this bill has been triple referred to the Assembly committees on Privacy and Consumer Protection, and Judiciary, upon passage out of this Committee.
As the use of AI in the workplace grows, it is being used in new and invasive ways that allow employers to collect an extensive array of data about workers. Data collection technologies available today include sensors embedded in workplace equipment, devices, and wearables that can capture data on worker location, activities, and interactions with coworkers; systems that log keystrokes and capture screenshots; GPS technologies embedded in vehicles; and more.
Increasingly, companies are using worker-generated data to train AI systems designed to automate tasks and streamline operations. In industries ranging from logistics and finance to customer service, employees are being asked to document workflows, record interactions, or label data, which is then used to automate their roles.
For example, Google reorganized its ad sales division and cut jobs after employees trained the AI product "Gemini for Sales." Amazon similarly used driver data to train autonomous routing systems. JPMorgan Chase implemented AI for contract analysis, which is trained by legal analysts.
The author provides the following additional examples:
- “Call center workers have felt the brunt of AI replacement and forced training for years. Since 2023, AT&T customer services workers have experienced AI deployment impacting their scope of work as AI agents have become integrated into their conversations with customers. By listening to customer conversations AI agents have become capable of drafting customer service scripts and even answering calls to solve simple customer questions. As time progresses and the system learns from future human interactions, call center workers will become increasingly vulnerable to deskilling and displacement.”
- “Software engineers have similarly faced job loss and hiring reductions due to increasingly sophisticated AI systems. The job market has plummeted for software engineers as many remaining engineers are forced to work alongside and even train AI models capable of improving at notable rates. In October 2025, reports demonstrated that the tech sector wiped out 33,281 jobs due to AI integration.”
- “Mercor, an agentic AI developer, trains their systems by hiring independent contractors to train AI agents in medicine, software engineering, law, journalism, psychology, finance, and public policy.”
Author’s Statements:
According to the author, “Data powers today’s artificial intelligence economy, and workers are increasingly the source of that data. As employers adopt AI tools, worker data is being captured, analyzed, and in some cases sold. This creates a troubling reality where workers may be unknowingly contributing to technologies that could replace them.
While using data to improve efficiency is not new, the scale and speed of AI today are unprecedented. Workers across industries are already facing job displacement, deskilling, and fewer opportunities as AI systems advance. At the same time, worker data has become a valuable commodity in a rapidly growing data economy, with little transparency or protection for those generating it.
AB 2027 sets clear boundaries on the use of worker data. It prohibits employers from using worker data to train AI systems that automate or replace jobs and prevents the sale or sharing of that data for that purpose. It also ensures that technology developers cannot use worker data collected through their tools to train their own systems or third-party AI meant to replace workers.
AB 2027 is about fairness, dignity, and economic security. Workers should not be used to train their own replacements. This bill takes an important step to protect workers and ensure AI is developed responsibly.”
Per HR 39 (Gipson, 2021), the author adds that the measure supports equity in that “the danger of job loss due to automation impacts every industry and worker across the state. Following historical trends, however, low income and marginalized communities that do not have adequate workplace representation are likely to suffer immense consequences far greater than Californians who have access to wealth, representation, and have the resources necessary to temporarily pivot careers or financially cover their loss of income. AB 2027 prevents workers from these very underserved communities from being used to further exacerbate their own incoming challenges of forced AI training and automation.”
Committee Comments:
Given the protections the Legislature has extended to consumers with regards to their data under the CCPA and CPRA, it is logical that the Legislature might establish guardrails for employers’ use of worker data. However, as the opposition points out, there are questions as to the practicality of implementation and enforcement of this bill as currently written. More specificity and clearer definitions are likely needed, as well as narrower prohibitions to ensure that business operations are not unduly hindered or workers disadvantaged.
Amendments in the future:
The author has conveyed to the Committee that he intends to make the following amendments subsequent to this Committee:
- 1572(e)(1): Except as provided in paragraph (2), in addition to any other remedy, an employer who violates this part may be subject to a penalty of up to five hundred dollars ($500)
per workerfor each violation. - Section 1572(f): An action brought pursuant to this section may be brought in the superior court in any county in which the violation in question is alleged to have occurred or in which the employer resides
or transacts business.
Arguments in Support
A coalition of labor and worker organizations, including the California Federation of Labor Unions, writes in support that “Data used to train large language models powers the artificial intelligence economy. As tech companies develop and sell enterprise AI systems to employers, worker data is a critical resource. To mine that data, tech companies are selling employers AI products designed to maximize the extraction of worker data and employers have found a lucrative market selling worker data to brokers.”
“…The strip mining of worker data is not simply a privacy or worker dignity issue; the scope and scale of AI surveillance and task automation threaten mass job loss across all industries. The development and deployment of AI tools to potentially replace or erode human jobs is happening at a velocity that makes it difficult for society to adapt or even mitigate the damage to workers, families, the economy, and society. Though tech companies try to soothe anxiety with the promise of new job creation, there is little evidence that those jobs will appear in time, or at all, to mitigate the mass job loss and economic disruption AI may cause.”
Arguments in Opposition
A coalition of business organizations, including the California Chamber of Commerce, state in opposition that “AB 2027 would impose sweeping and unworkable restrictions on the use of worker data that will significantly hinder responsible innovation and expose employers to substantial litigation risk. AB 2027 would undermine the ability of businesses to use common technologies that improve workplace efficiency, safety, and compliance with existing employment laws.
AB 2027 would broadly prohibit employers and vendors from using worker data to develop or deploy artificial intelligence (AI) systems that could replicate or automate job functions and create joint liability between employers and vendors for violations. AB 2027’s overly broad expansive definitions, strict liability framework, and private right of action will discourage responsible use of technology and place California employers at a competitive disadvantage.”
Prior and Related Legislation
AB 1883 (Bryan) of 2026 would prohibit an employer from using certain types of workplace surveillance tools, or using workplace surveillance tools to violate or prevent compliance with laws, or infer information about a worker’s legally-protected status or activities. Provides for a civil penalty, enforcement by the LC or a public prosecutor, and a private right of action. Pending in the Assembly Privacy and Consumer Protection Committee.
AB 1898 (Schultz) of 2026 would, among other things, require an employer to provide a written notice to an employee that a workplace AI tool, as defined, was used to assist the employer in making employment-related decisions or to surveil the workplace. Pending in the Assembly Judiciary Committee.
SB 951 (Reyes) of 2026 would establish the California Worker Technological Displacement Act, which would require a covered employer, as defined, to provide at least a 90-day advanced written notice, as described, before any technological displacement or termination of contract affecting 25 or more workers or 25 percent of the workforce, whichever is less; would require a covered employer to provide that notice to affected employees, the EDD, and specified state and local entities. Pending in the Senate Labor, Public Employment and Retirement Committee.
SB 947 (McNerney) of 2026 would, among other things, prohibit an employer from using an automated decision system (ADS) to perform certain functions and would limit the purposes for and way in which an ADS may be used. The bill would authorize a worker to request, and require an employer to provide, a copy of the most recent 12 months of the worker’s own data primarily used by an ADS to make a disciplinary, termination, or deactivation decision, as specified. The bill would require an employer that uses an ADS to assist in making a disciplinary, termination, or deactivation decision to provide the affected worker with a written post-use notice, as specified. Pending in the Senate Labor, Public Employment and Retirement Committee.
AB 1331 (Elhawary) of 2025 would limit the use of workplace surveillance tools, as defined, by employers, including by prohibiting an employer from monitoring or surveilling workers in employee-only, employer-designated areas, as specified. On the Senate Inactive File.
AB 1221 (Bryan) of 2025, among other things, would have 1) required an employer, at least 30 days before introducing a workplace surveillance tool, as defined, to provide a worker who will be affected a written notice that includes specified disclosures; 2) created certain prohibitions and requirements related to the use of worker data, as defined; 3) prohibited employers from using workplace surveillance tools in specified ways; and 4) provided for enforcement by the LC, employees and representatives, and public prosecutors, and subjected employers in violation to a civil penalty. Held in the Assembly Appropriations Committee.
SB 7 (McNerney) of 2025 would have regulated the use of ADS in the employment setting. Among other things, this bill would have 1) required an employer to provide a written notice that an ADS is in use at the workplace to all workers that will foreseeably be directly affected by the ADS; 2) prohibited in some instances and in others limits the use of an ADS by an employer; 3) provided worker anti-retaliation protections for exercising their rights under these provisions; and 4) specified enforcement provisions that include penalties and relief for violations. Vetoed by Governor Newsom.
SB 238 (Smallwood-Cuevas) of 2025 would require an employer to annually provide a notice to the DIR of all the workplace surveillance tools the employer is using in the workplace. The bill would require the notice to include, among other information, the personal information that will be collected from workers and consumers and whether they will have the option of opting out of the collection of personal information. Pending in the Assembly Consumer Privacy and Protection Committee.
AB 1651 (Kalra) of 2022 would have established limitations on the use of data-driven technologies in the workplace by requiring employers to notify workers prior to data collection, initiating electronic monitoring, and deploying algorithms; as well as would have required the technology be used pursuant to a valid business practice and be job-related, and that employers conduct impact assessments with worker input for algorithms. Died in the Assembly Privacy and Consumer Protection Committee.
REGISTERED SUPPORT / OPPOSITION:
Support
California Federation of Labor Unions, AFL-CIO (Sponsor)
Alameda Labor Council
California Alliance for Retired Americans
California Faculty Association
California Federation of Teachers AFL-CIO
California School Employees Association
California State Legislative Board of the Sheet Metal, Air, Rail and Transportation Workers - Transportation Division (SMART-TD)
California Teachers Association
Central Coast Labor Council
Central Labor Council, Fresno-Madera-Tulare-Kings Counties, AFL-CIO
Inland Empire Labor Council, AFL-CIO
North Bay Labor Council
North Valley Labor Federation
Oakland Privacy
Orange County Labor Federation, AFL-CIO
San Mateo County Central Labor Council
TechEquity Action
UAW Region 6
Opposition
Acclamation Insurance Management Services
Allied Managed Care
Association of California Healthcare Districts
California Apartment Association
California Chamber of Commerce
California Farm Bureau
California Hospital Association
California Manufacturers & Technology Association
California Retailers Association
California Special Districts Association
California State Association of Counties
California Trucking Association
Cinema Association of California
Civil Justice Association of California
Flasher Barricade Association
Leading Age California
League of California Cities
Los Angeles Area Chamber of Commerce
Rural County Representatives of California
Technet
Urban Counties of California
Analysis Prepared by: Erin Hickey / L. & E. /
Discussed in Hearing
Assembly Standing Committee on Privacy and Consumer Protection
Assembly Standing Committee on Labor and Employment
Bill Author
Bill Co-Author(s):