SB 1000: California AI Transparency Act.
- Session Year: 2025-2026
- House: Senate
- Latest Version Date: 2026-09-30
Current Status:
Passed
(2026-09-30: Chaptered by Secretary of State. Chapter 861, Statutes of 2026.)
Introduced
In Committee
Awaiting First Chamber Vote
Passed First Chamber
In Committee
Awaiting Second Chamber Vote
Passed Second Chamber
Enacted
Existing law, the California AI Transparency Act (CATA), beginning August 2, 2026, generally regulates provenance data disclosure in content generated by artificial intelligence (AI), including by requiring a covered provider to make available an AI detection tool at no cost to the user that meets certain criteria. Existing law requires a covered provider to offer the user the option to include a certain manifest disclosure in image, video, or audio content, or content that is any combination thereof, created or altered by the covered providers generative artificial intelligence (GenAI) system and requires a covered provider to include a certain latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered providers GenAI system. Existing law defines covered provider for these purposes to mean a person that creates, codes, or otherwise produces a generative AI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state.
This bill would recast those provisions to, among other changes, delete the user threshold from the definition of covered provider, replace the term AI detection tool with disclosure verification tool, delete the above-described requirement of a covered provider to offer the user the option to include a manifest disclosure in content, and additionally require a covered provider to include in the above-described latent disclosure whether the GenAI system created or altered the content. The bill would delay CATAs operation with respect to a GenAI system that is designed to primarily function as assistive technology, as defined, would prohibit a covered provider from falsely representing that a GenAI system is designed to primarily function as assistive technology, and would punish a covered provider who makes such a false representation with a certain civil action, as specified.
This bill would declare that it is to take effect immediately as an urgency statute.
Existing law, the California AI Transparency Act (CATA), beginning August 2, 2026, generally regulates provenance data disclosure in content generated by artificial intelligence (AI), including by requiring a covered provider to make available an AI detection tool at no cost to the user that meets certain criteria. Existing law requires a covered provider to offer the user the option to include a certain manifest disclosure in image, video, or audio content, or content that is any combination thereof, created or altered by the covered providers generative artificial intelligence (GenAI) system and requires a covered provider to include a certain latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered providers GenAI system. Existing law defines covered provider for these purposes to mean a person that creates, codes, or otherwise produces a generative AI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state.
This bill would recast those provisions to, among other changes, delete the user threshold from the definition of covered provider, replace the term AI detection tool with disclosure verification tool, delete the above-described requirement of a covered provider to offer the user the option to include a manifest disclosure in content, and additionally require a covered provider to include in the above-described latent disclosure whether the GenAI system created or altered the content. The bill would delay CATAs operation with respect to a GenAI system that is designed to primarily function as assistive technology, as defined, would prohibit a covered provider from falsely representing that a GenAI system is designed to primarily function as assistive technology, and would punish a covered provider who makes such a false representation with a certain civil action, as specified.
This bill would declare that it is to take effect immediately as an urgency statute.
Existing law, the California AI Transparency Act, Act (CATA), beginning August 2, 2026, generally regulates provenance data disclosure in content generated by artificial intelligence (AI), including by requiring a covered provider to make available an AI detection tool at no cost to the user that meets certain criteria. Existing law requires a covered provider to offer the user the option to include a certain manifest disclosure in image, video, or audio content, or content that is any combination thereof, created or altered by the covered providers generative artificial intelligence (GenAI) system and requires a covered provider to include a certain latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered providers GenAI system. Existing law defines covered provider for these purposes to mean a person that creates, codes, or otherwise produces a generative AI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state.
This bill would recast those provisions to, among other changes, delete the user threshold from the definition of covered provider, replace the term AI detection tool with disclosure verification tool, delete the above-described requirement of a covered provider to offer the user the option to include a manifest disclosure in content, and additionally require a covered provider to include in the above-described latent disclosure whether the GenAI system created or altered the content. The bill would delay CATAs operation with respect to a GenAI system that is designed to primarily function as assistive technology, as defined, would prohibit a covered provider from falsely representing that a GenAI system is designed to primarily function as assistive technology, and would punish a covered provider who makes such a false representation with a certain civil action, as specified.
This bill would declare that it is to take effect immediately as an urgency statute.
Existing law, the California AI Transparency Act, beginning August 2, 2026, generally regulates provenance data disclosure in content generated by artificial intelligence (AI), including by requiring a covered provider to make available an AI detection tool at no cost to the user that meets certain criteria. Existing law requires a covered provider to offer the user the option to include a certain manifest disclosure in image, video, or audio content, or content that is any combination thereof, created or altered by the covered providers generative artificial intelligence (GenAI) system and requires a covered provider to include a certain latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered providers GenAI system. Existing law defines covered provider for these purposes to mean a person that creates, codes, or otherwise produces a generative AI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state.
This bill would recast those provisions to, among other changes, delete the user threshold from the definition of covered provider, replace the term AI detection tool with disclosure verification tool, delete the above-described requirement of a covered provider to offer the user the option to include a manifest disclosure in content, and additionally require a covered provider to include in the above-described latent disclosure whether the content is generated or modified by artificial intelligence. GenAI system created or altered the content.
This bill would declare that it is to take effect immediately as an urgency statute.
Existing law, the California AI Transparency Act, beginning August 2, 2026, generally regulates provenance data disclosure in content generated by artificial intelligence (AI), including by requiring a covered provider to make available an AI detection tool at no cost to the user that meets certain criteria. Existing law requires a covered provider to offer the user the option to include a certain manifest disclosure in image, video, or audio content, or content that is any combination thereof, created or altered by the covered providers generative artificial intelligence (GenAI) system and requires a covered provider to include a certain latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered providers GenAI system. Existing law defines covered provider for these purposes to mean a person that creates, codes, or otherwise produces a generative AI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state.
This bill would recast those provisions to, among other changes, delete the user threshold from the definition of covered provider, replace the term AI detection tool with disclosure verification tool, delete the above-described requirement of a covered provider to offer the user the option to include a manifest disclosure in content, and additionally require a covered provider to include in the above-described latent disclosure whether the content is generated or modified by artificial intelligence.
This bill would also exempt from those provisions a product, service, or application that is primarily marketed and intended to facilitate accessibility for disabled individuals and is designed to prevent content generated by the product, service, internet website, or application from being downloaded, stored, or otherwise appropriated, to the extent technically feasible.This bill would declare that it is to take effect immediately as an urgency statute.
Existing law, the California AI Transparency Act, beginning August 2, 2026, generally regulates provenance data disclosure in content generated by artificial intelligence (AI), including by requiring a covered provider to make available an AI detection tool at no cost to the user that meets certain criteria. Existing law requires a covered provider to offer the user the option to include a certain manifest disclosure in image, video, or audio content, or content that is any combination thereof, created or altered by the covered providers generative artificial intelligence (GenAI) system and requires a covered provider to include a certain latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered providers GenAI system. Existing law defines covered provider for these purposes to mean a person that creates, codes, or otherwise produces a generative AI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state.
This bill would recast those provisions to, among other changes, delete the user threshold from the definition of covered provider, replace the term AI detection tool with disclosure verification tool, delete the above-described requirement of a covered provider to offer the user the option to include a manifest disclosure in content, and additionally require a covered provider to include in the above-described latent disclosure whether the content is generated or modified by artificial intelligence.
This bill would also exempt from those provisions a product, service, or application that is primarily marketed and intended to facilitate accessibility for disabled individuals and is designed to prevent content generated by the product, service, internet website, or application from being downloaded, stored, or otherwise appropriated, to the extent technically feasible.
This bill would declare that it is to take effect immediately as an urgency statute.
The Legislature finds and declares all of the following:
(a)Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence (AI) system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative AI to create content for other malicious purposes.
(b)Conveying conflicting manifest and latent provenance information to consumers is detrimental to trust in both the provenance of content and the authenticity of any person, entity, or institution conveyed in that content.
(c)AI provenance technology is still developing, and ensuring that information conveyed to consumers regarding a piece of contents provenance is accurate, accessible, and resilient to common transformations to the extent technically feasible over time requires legislation to be adaptive and technologically agnostic.
(d)Accordingly, encouraging the adoption, interoperability, and continued development of widely accepted technical standards is necessary to promote reliable, scalable, and evolvable content provenance mechanisms.
SEC. 2.Section 22757.1 of the Business and Professions Code is amended to read:22757.1.As used in this chapter:
(a)Artificial intelligence or AI 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)Assistive technology means an item, piece of equipment, or product system, whether acquired commercially, modified, or customized, that is used to increase, maintain, or improve functional capabilities of individuals with disabilities and any service that directly assists an individual with a disability in the selection, acquisition, or use of the item, equipment, or product system.
(c)Capture device means a device that can record photographs, audio, or video content, including, but not limited to, video and still photography cameras, mobile phones with built-in cameras or microphones, and voice recorders.
(d)(1)Capture device manufacturer means a person who produces a capture device for sale in the state.
(2)Capture device manufacturer does not include a person exclusively engaged in the assembly of a capture device.
(e)Covered provider means a person that creates, codes, or otherwise produces a generative artificial intelligence system that is publicly accessible within the geographic boundaries of the state.
(f)Digital signature means a cryptography-based method that identifies the user or entity that attests to the information provided in the signed section.
(g)Generative artificial intelligence system or GenAI system means an artificial intelligence that can generate derived synthetic content, including text, images, video, and audio, that emulates the structure and characteristics of the systems training data.
(h)GenAI hosting platform means an internet website or application that makes available for download the source code or model weights a generative artificial intelligence system by a resident of the state, regardless of whether the terms of that use include compensation.
(i)(1)Large online platform means a public-facing social media platform, file-sharing platform, mass messaging platform, or stand-alone search engine that distributes content to users who did not create or collaborate in creating the content that exceeded 2,000,000 unique monthly users during the preceding 12 months.
(2)Large online platform does not include either of the following:
(A)A broadband internet access service, as defined in Section 3100 of the Civil Code.
(B)A telecommunications service, as defined in Section 153 of Title 47 of the United States Code.
(j)Mass messaging platform means a direct messaging platform that allows users to distribute content to more than 100 users simultaneously.
(k)Metadata means structural or descriptive information about data.
(l)Minor modification means any of the following alterations:
(1)A change to brightness, contrast, or color.
(2)Sharpening.
(3)Saturating.
(4)File resizing.
(5)Scaling.
(6)Cropping.
(7)File format conversions.
(8)Denoising and removal of background noise in audio.
(m)Personal information has the same meaning as defined in Section 1798.140 of the Civil Code.
(n)Provenance data means information about the origin of a piece of digital content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
(o)System provenance data means provenance data that is not reasonably capable of being associated with a particular user and that contains either of the following:
(1)Information regarding the type of device, system, or service that was used to generate a piece of digital content.
(2)Information related to content authenticity.
SEC. 3.Section 22757.2 of the Business and Professions Code is amended to read:22757.2.(a)A covered provider shall make available a disclosure verification tool at no cost to the user that meets all of the following criteria:
(1)The tool allows a user to assess whether image, video, or audio content, or content that is any combination thereof, was created or altered, except by minor modification, by the covered providers GenAI system.
(2)The tool outputs any system provenance data that is detected in the content.
(3)(A)Except as provided in subparagraph (B), the tool does not output any personal information that is detected in the content.
(B)The tool may output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content pursuant to a notice that does both of the following:
(i)Informs the user of the personal information that may be output by the tool.
(ii)Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the files permanent digital footprint and cannot be retracted from copies already in circulation.
(4)(A)Subject to subparagraph (B), the tool is publicly accessible.
(B)A covered provider may impose reasonable limitations on access to the tool to prevent, or respond to, demonstrable risks to the security or integrity of its GenAI system or to prevent misuse of the tool for malicious purposes.
(5)The tool allows a user to upload content or provide a uniform resource locator (URL) linking to online content.
(6)The tool supports technology, including an application programming interface, that allows a user to invoke the tool without visiting the covered providers internet website.
(b)A covered provider shall collect user feedback related to the efficacy of the covered providers AI disclosure verification tool and incorporate relevant feedback into any attempt to improve the efficacy of the tool.
(c)(1)Except as provided in paragraph (2), a covered provider shall not collect, use, retain, sell, share, or otherwise make available personal information derived either from a user of the covered providers AI disclosure verification tool or from any content processed by the disclosure verification tool beyond what is strictly necessary to comply with this chapter.
(2)A covered provider may collect, retain, or use personal information for the sole purpose of communicating with a user that opts into being contacted by the covered provider.
(d)A covered provider shall not make access to the covered providers GenAI system or disclosure verification tool contingent upon a user providing personal information beyond what is strictly necessary to comply with this chapter.
(e)A covered provider may satisfy the requirements of this section by directing a user to a third-party disclosure verification tool that is all of the following:
(1)In compliance with this section.
(2)Compatible with latent disclosures included in content produced or altered by the covered providers GenAI system.
(3)Clearly and conspicuously accessible through the user interface of the covered providers GenAI system.
SEC. 4.Section 22757.3 of the Business and Professions Code is amended to read:22757.3.(a)To the extent it is technically feasible, a covered provider shall include a latent disclosure that meets all of the following criteria in any image, video, or audio content, or content that is any combination thereof, created or altered, except by minor modification, by the covered providers GenAI system:
(1)The disclosure conveys all of the following information, either directly or through a link to a permanent internet website:
(A)The name of the covered provider.
(B)The name and version information of the GenAI system that created or altered the content.
(C)The time and date of the contents creation or alteration.
(D)A unique identifier.
(E)Whether the GenAI system created or altered the content.
(F)On and after January 1, 2029, whether the GenAI system is designed to primarily function as assistive technology.
(2)The disclosure is permanent or extraordinarily difficult to remove or tamper with.
(3)The disclosure is compatible with the covered providers disclosure verification tool.
(4)The disclosure is compliant or interoperable with widely recognized industry standards.
(b)(1)If a covered provider licenses its GenAI system to a third party, the covered provider shall, when it provides the license, notify the licensee of the licensees obligations under this chapter.
(2)If a covered provider knows that an identifiable third-party licensee modified a licensed GenAI system such that it no longer complies with this chapter, the covered provider shall do either of the following:
(A)Terminate the licensees authorization to use the GenAI system within 72 hours of discovering the licensees action.
(B)Notify the licensee of both of the following within 72 hours of discovering the licensees action:
(i)The licensees noncompliance.
(ii)The requirement to report noncompliance with paragraph (3) to the Attorney General pursuant to paragraph (4).
(3)A third-party licensee that receives a notice under paragraph (2) shall modify the licensed GenAI system to be in compliance with this chapter or cease using or making available the licensed GenAI system, including a copy or modified version of the GenAI system and shall report which of those actions it took to the covered provider within 96 hours of receipt of the notice.
(4)If a covered provider does not receive a report pursuant to paragraph (3) or if a covered provider receives a report from a third-party licensee indicating that the third-party licensee did not, pursuant to paragraph (3), modify or cease using or making available the licensed GenAI system, the covered provider shall report the third-party licensees noncompliance with paragraph (3) to the Attorney General.
(5)This subdivision does not require a covered provider to monitor, investigate, or otherwise inquire into a third-party licensees use or modification of a licensed GenAI system.
(c)The Attorney General shall establish a mechanism to receive reports from covered providers submitted to the Attorney General pursuant to this section.
(d)A covered provider shall not falsely represent that a GenAI system is designed to primarily function as assistive technology.
SEC. 5.Section 22757.4 of the Business and Professions Code is amended to read:22757.4.(a)(1)A violator of this chapter shall be liable for a civil penalty in the amount of five thousand dollars ($5,000) per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel.
(2)A prevailing plaintiff in an action brought pursuant to this subdivision shall be entitled to all reasonable attorneys costs and fees.
(b)Each day that a covered provider, large online platform, or capture device manufacturer is in violation of this chapter shall be deemed a discrete violation.
(c)A civil action brought under this section before the effective date of the act adding this subdivision shall not be maintained if the alleged conduct does not violate this chapter on and after the effective date of the act adding this subdivision.
(d)This section does not apply with respect to a violation of subdivision (d) of Section 22757.3.
SEC. 6.Section 22757.4.1 is added to the Business and Professions Code, to read:22757.4.1.(a)(1)A violator of subdivision (d) of Section 22757.3 shall be liable for a civil penalty in the amount of fifty thousand dollars ($50,000) per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel.
(2)A prevailing plaintiff in an action brought pursuant to this subdivision shall be entitled to all reasonable attorneys costs and fees.
(b)Each day that a covered provider is in violation of this section shall be deemed a discrete violation.
(c)This section shall remain in effect until January 1, 2029, and as of that date is repealed.
SEC. 7.Section 22757.5 of the Business and Professions Code is amended to read:22757.5.(a)This chapter does not apply to any product, service, internet website, or application that provides exclusively non-user-generated videogames.
(b)Before January 1, 2029, this chapter does not apply to a GenAI system that is designed to primarily function as assistive technology.
SEC. 8.This act is an urgency statute necessary for the immediate preservation of the public peace, health, or safety within the meaning of Article IV of the California Constitution and shall go into immediate effect. The facts constituting the necessity are:
Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative artificial intelligence to create content for other malicious purposes, and accordingly, it is necessary that this act go into immediate effect.
The Legislature finds and declares all of the following:
(a)Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence (AI) system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative AI to create content for other malicious purposes.
(b)Conveying conflicting manifest and latent provenance information to consumers is detrimental to trust in both the provenance of content and the authenticity of any person, entity, or institution conveyed in that content.
(c)AI provenance technology is still developing, and ensuring that information conveyed to consumers regarding a piece of contents provenance is accurate, accessible, and resilient to common transformations to the extent technically feasible over time requires legislation to be adaptive and technologically agnostic.
(d)Accordingly, encouraging the adoption, interoperability, and continued development of widely accepted technical standards is necessary to promote reliable, scalable, and evolvable content provenance mechanisms.
SEC. 2.Section 22757.1 of the Business and Professions Code is amended to read:22757.1.As used in this chapter:
(a)Artificial intelligence or AI 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)Assistive technology means an item, piece of equipment, or product system, whether acquired commercially, modified, or customized, that is used to increase, maintain, or improve functional capabilities of individuals with disabilities and any service that directly assists an individual with a disability in the selection, acquisition, or use of the item, equipment, or product system.
(c)Capture device means a device that can record photographs, audio, or video content, including, but not limited to, video and still photography cameras, mobile phones with built-in cameras or microphones, and voice recorders.
(d)(1)Capture device manufacturer means a person who produces a capture device for sale in the state.
(2)Capture device manufacturer does not include a person exclusively engaged in the assembly of a capture device.
(e)Covered provider means a person that creates, codes, or otherwise produces a generative artificial intelligence system that is publicly accessible within the geographic boundaries of the state.
(f)Digital signature means a cryptography-based method that identifies the user or entity that attests to the information provided in the signed section.
(g)Generative artificial intelligence system or GenAI system means an artificial intelligence that can generate derived synthetic content, including text, images, video, and audio, that emulates the structure and characteristics of the systems training data.
(h)GenAI hosting platform means an internet website or application that makes available for download the source code or model weights a generative artificial intelligence system by a resident of the state, regardless of whether the terms of that use include compensation.
(i)(1)Large online platform means a public-facing social media platform, file-sharing platform, mass messaging platform, or stand-alone search engine that distributes content to users who did not create or collaborate in creating the content that exceeded 2,000,000 unique monthly users during the preceding 12 months.
(2)Large online platform does not include either of the following:
(A)A broadband internet access service, as defined in Section 3100 of the Civil Code.
(B)A telecommunications service, as defined in Section 153 of Title 47 of the United States Code.
(j)Mass messaging platform means a direct messaging platform that allows users to distribute content to more than 100 users simultaneously.
(k)Metadata means structural or descriptive information about data.
(l)Minor modification means any of the following alterations:
(1)A change to brightness, contrast, or color.
(2)Sharpening.
(3)Saturating.
(4)File resizing.
(5)Scaling.
(6)Cropping.
(7)File format conversions.
(8)Denoising and removal of background noise in audio.
(m)Personal information has the same meaning as defined in Section 1798.140 of the Civil Code.
(n)Provenance data means information about the origin of a piece of digital content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
(o)System provenance data means provenance data that is not reasonably capable of being associated with a particular user and that contains either of the following:
(1)Information regarding the type of device, system, or service that was used to generate a piece of digital content.
(2)Information related to content authenticity.
SEC. 3.Section 22757.2 of the Business and Professions Code is amended to read:22757.2.(a)A covered provider shall make available a disclosure verification tool at no cost to the user that meets all of the following criteria:
(1)The tool allows a user to assess whether image, video, or audio content, or content that is any combination thereof, was created or altered, except by minor modification, by the covered providers GenAI system.
(2)The tool outputs any system provenance data that is detected in the content.
(3)(A)Except as provided in subparagraph (B), the tool does not output any personal information that is detected in the content.
(B)The tool may output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content pursuant to a notice that does both of the following:
(i)Informs the user of the personal information that may be output by the tool.
(ii)Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the files permanent digital footprint and cannot be retracted from copies already in circulation.
(4)(A)Subject to subparagraph (B), the tool is publicly accessible.
(B)A covered provider may impose reasonable limitations on access to the tool to prevent, or respond to, demonstrable risks to the security or integrity of its GenAI system or to prevent misuse of the tool for malicious purposes.
(5)The tool allows a user to upload content or provide a uniform resource locator (URL) linking to online content.
(6)The tool supports technology, including an application programming interface, that allows a user to invoke the tool without visiting the covered providers internet website.
(b)A covered provider shall collect user feedback related to the efficacy of the covered providers AI disclosure verification tool and incorporate relevant feedback into any attempt to improve the efficacy of the tool.
(c)(1)Except as provided in paragraph (2), a covered provider shall not collect, use, retain, sell, share, or otherwise make available personal information derived either from a user of the covered providers AI disclosure verification tool or from any content processed by the disclosure verification tool beyond what is strictly necessary to comply with this chapter.
(2)A covered provider may collect, retain, or use personal information for the sole purpose of communicating with a user that opts into being contacted by the covered provider.
(d)A covered provider shall not make access to the covered providers GenAI system or disclosure verification tool contingent upon a user providing personal information beyond what is strictly necessary to comply with this chapter.
(e)A covered provider may satisfy the requirements of this section by directing a user to a third-party disclosure verification tool that is all of the following:
(1)In compliance with this section.
(2)Compatible with latent disclosures included in content produced or altered by the covered providers GenAI system.
(3)Clearly and conspicuously accessible through the user interface of the covered providers GenAI system.
SEC. 4.Section 22757.3 of the Business and Professions Code is amended to read:22757.3.(a)To the extent it is technically feasible, a covered provider shall include a latent disclosure that meets all of the following criteria in any image, video, or audio content, or content that is any combination thereof, created or altered, except by minor modification, by the covered providers GenAI system:
(1)The disclosure conveys all of the following information, either directly or through a link to a permanent internet website:
(A)The name of the covered provider.
(B)The name and version information of the GenAI system that created or altered the content.
(C)The time and date of the contents creation or alteration.
(D)A unique identifier.
(E)Whether the GenAI system created or altered the content.
(F)On and after January 1, 2029, whether the GenAI system is designed to primarily function as assistive technology.
(2)The disclosure is permanent or extraordinarily difficult to remove or tamper with.
(3)The disclosure is compatible with the covered providers disclosure verification tool.
(4)The disclosure is compliant or interoperable with widely recognized industry standards.
(b)(1)If a covered provider licenses its GenAI system to a third party, the covered provider shall, when it provides the license, notify the licensee of the licensees obligations under this chapter.
(2)If a covered provider knows that an identifiable third-party licensee modified a licensed GenAI system such that it no longer complies with this chapter, the covered provider shall do either of the following:
(A)Terminate the licensees authorization to use the GenAI system within 72 hours of discovering the licensees action.
(B)Notify the licensee of both of the following within 72 hours of discovering the licensees action:
(i)The licensees noncompliance.
(ii)The requirement to report noncompliance with paragraph (3) to the Attorney General pursuant to paragraph (4).
(3)A third-party licensee that receives a notice under paragraph (2) shall modify the licensed GenAI system to be in compliance with this chapter or cease using or making available the licensed GenAI system, including a copy or modified version of the GenAI system and shall report which of those actions it took to the covered provider within 96 hours of receipt of the notice.
(4)If a covered provider does not receive a report pursuant to paragraph (3) or if a covered provider receives a report from a third-party licensee indicating that the third-party licensee did not, pursuant to paragraph (3), modify or cease using or making available the licensed GenAI system, the covered provider shall report the third-party licensees noncompliance with paragraph (3) to the Attorney General.
(5)This subdivision does not require a covered provider to monitor, investigate, or otherwise inquire into a third-party licensees use or modification of a licensed GenAI system.
(c)The Attorney General shall establish a mechanism to receive reports from covered providers submitted to the Attorney General pursuant to this section.
(d)A covered provider shall not falsely represent that a GenAI system is designed to primarily function as assistive technology.
SEC. 5.Section 22757.4 of the Business and Professions Code is amended to read:22757.4.(a)(1)A violator of this chapter shall be liable for a civil penalty in the amount of five thousand dollars ($5,000) per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel.
(2)A prevailing plaintiff in an action brought pursuant to this subdivision shall be entitled to all reasonable attorneys costs and fees.
(b)Each day that a covered provider, large online platform, or capture device manufacturer is in violation of this chapter shall be deemed a discrete violation.
(c)A civil action brought under this section before the effective date of the act adding this subdivision shall not be maintained if the alleged conduct does not violate this chapter on and after the effective date of the act adding this subdivision.
(d)This section does not apply with respect to a violation of subdivision (d) of Section 22757.3.
SEC. 6.Section 22757.4.1 is added to the Business and Professions Code, to read:22757.4.1.(a)(1)A violator of subdivision (d) of Section 22757.3 shall be liable for a civil penalty in the amount of fifty thousand dollars ($50,000) per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel.
(2)A prevailing plaintiff in an action brought pursuant to this subdivision shall be entitled to all reasonable attorneys costs and fees.
(b)Each day that a covered provider is in violation of this section shall be deemed a discrete violation.
(c)This section shall remain in effect until January 1, 2029, and as of that date is repealed.
SEC. 7.Section 22757.5 of the Business and Professions Code is amended to read:22757.5.(a)This chapter does not apply to any product, service, internet website, or application that provides exclusively non-user-generated videogames.
(b)Before January 1, 2029, this chapter does not apply to a GenAI system that is designed to primarily function as assistive technology.
SEC. 8.This act is an urgency statute necessary for the immediate preservation of the public peace, health, or safety within the meaning of Article IV of the California Constitution and shall go into immediate effect. The facts constituting the necessity are:
Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative artificial intelligence to create content for other malicious purposes, and accordingly, it is necessary that this act go into immediate effect.
The Legislature finds and declares all of the following:
(a)Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence (AI) system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative AI to create content for other malicious purposes.
(b)Conveying conflicting manifest and latent provenance information to consumers is detrimental to trust in both the provenance of content and the authenticity of any person, entity, or institution conveyed in that content.
(c)AI provenance technology is still developing, and ensuring that information conveyed to consumers regarding a piece of contents provenance is accurate, accessible, and resilient to common transformations to the extent technically feasible over time requires legislation to be adaptive and technologically agnostic.
(d)Accordingly, encouraging the adoption, interoperability, and continued development of widely accepted technical standards is necessary to promote reliable, scalable, and evolvable content provenance mechanisms.
SEC. 2.Section 22757.1 of the Business and Professions Code is amended to read:22757.1.As used in this chapter:
(a)Artificial intelligence or AI 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)Assistive technology means an item, piece of equipment, or product system, whether acquired commercially, modified, or customized, that is used to increase, maintain, or improve functional capabilities of individuals with disabilities and any service that directly assists an individual with a disability in the selection, acquisition, or use of the item, equipment, or product system.
(c)Capture device means a device that can record photographs, audio, or video content, including, but not limited to, video and still photography cameras, mobile phones with built-in cameras or microphones, and voice recorders.
(d)(1)Capture device manufacturer means a person who produces a capture device for sale in the state.
(2)Capture device manufacturer does not include a person exclusively engaged in the assembly of a capture device.
(e)Covered provider means a person that creates, codes, or otherwise produces a generative artificial intelligence system that is publicly accessible within the geographic boundaries of the state.
(f)Digital signature means a cryptography-based method that identifies the user or entity that attests to the information provided in the signed section.
(g)Generative artificial intelligence system or GenAI system means an artificial intelligence that can generate derived synthetic content, including text, images, video, and audio, that emulates the structure and characteristics of the systems training data.
(h)GenAI hosting platform means an internet website or application that makes available for download the source code or model weights a generative artificial intelligence system by a resident of the state, regardless of whether the terms of that use include compensation.
(i)(1)Large online platform means a public-facing social media platform, file-sharing platform, mass messaging platform, or stand-alone search engine that distributes content to users who did not create or collaborate in creating the content that exceeded 2,000,000 unique monthly users during the preceding 12 months.
(2)Large online platform does not include either of the following:
(A)A broadband internet access service, as defined in Section 3100 of the Civil Code.
(B)A telecommunications service, as defined in Section 153 of Title 47 of the United States Code.
(j)Mass messaging platform means a direct messaging platform that allows users to distribute content to more than 100 users simultaneously.
(k)Metadata means structural or descriptive information about data.
(l)Minor modification means any of the following alterations:
(1)A change to brightness, contrast, or color.
(2)Sharpening.
(3)Saturating.
(4)File resizing.
(5)Scaling.
(6)Cropping.
(7)File format conversions.
(8)Denoising and removal of background noise in audio.
(m)Personal information has the same meaning as defined in Section 1798.140 of the Civil Code.
(n)Provenance data means information about the origin of a piece of digital content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
(o)System provenance data means provenance data that is not reasonably capable of being associated with a particular user and that contains either of the following:
(1)Information regarding the type of device, system, or service that was used to generate a piece of digital content.
(2)Information related to content authenticity.
SEC. 3.Section 22757.2 of the Business and Professions Code is amended to read:22757.2.(a)A covered provider shall make available a disclosure verification tool at no cost to the user that meets all of the following criteria:
(1)The tool allows a user to assess whether image, video, or audio content, or content that is any combination thereof, was created or altered, except by minor modification, by the covered providers GenAI system.
(2)The tool outputs any system provenance data that is detected in the content.
(3)(A)Except as provided in subparagraph (B), the tool does not output any personal information that is detected in the content.
(B)The tool may output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content pursuant to a notice that does both of the following:
(i)Informs the user of the personal information that may be output by the tool.
(ii)Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the files permanent digital footprint and cannot be retracted from copies already in circulation.
(4)(A)Subject to subparagraph (B), the tool is publicly accessible.
(B)A covered provider may impose reasonable limitations on access to the tool to prevent, or respond to, demonstrable risks to the security or integrity of its GenAI system or to prevent misuse of the tool for malicious purposes.
(5)The tool allows a user to upload content or provide a uniform resource locator (URL) linking to online content.
(6)The tool supports technology, including an application programming interface, that allows a user to invoke the tool without visiting the covered providers internet website.
(b)A covered provider shall collect user feedback related to the efficacy of the covered providers AI disclosure verification tool and incorporate relevant feedback into any attempt to improve the efficacy of the tool.
(c)(1)Except as provided in paragraph (2), a covered provider shall not collect, use, retain, sell, share, or otherwise make available personal information derived either from a user of the covered providers AI disclosure verification tool or from any content processed by the disclosure verification tool beyond what is strictly necessary to comply with this chapter.
(2)A covered provider may collect, retain, or use personal information for the sole purpose of communicating with a user that opts into being contacted by the covered provider.
(d)A covered provider shall not make access to the covered providers GenAI system or disclosure verification tool contingent upon a user providing personal information beyond what is strictly necessary to comply with this chapter.
(e)A covered provider may satisfy the requirements of this section by directing a user to a third-party disclosure verification tool that is all of the following:
(1)In compliance with this section.
(2)Compatible with latent disclosures included in content produced or altered by the covered providers GenAI system.
(3)Clearly and conspicuously accessible through the user interface of the covered providers GenAI system.
SEC. 4.Section 22757.3 of the Business and Professions Code is amended to read:22757.3.(a)To the extent it is technically feasible, a covered provider shall include a latent disclosure that meets all of the following criteria in any image, video, or audio content, or content that is any combination thereof, created or altered, except by minor modification, by the covered providers GenAI system:
(1)The disclosure conveys all of the following information, either directly or through a link to a permanent internet website:
(A)The name of the covered provider.
(B)The name and version information of the GenAI system that created or altered the content.
(C)The time and date of the contents creation or alteration.
(D)A unique identifier.
(E)Whether the GenAI system created or altered the content.
(F)Whether On and after January 1, 2029, whether the GenAI system is designed to primarily function as assistive technology.
(2)The disclosure is permanent or extraordinarily difficult to remove or tamper with.
(3)The disclosure is compatible with the covered providers disclosure verification tool.
(4)The disclosure is compliant or interoperable with widely recognized industry standards.
(b)(1)If a covered provider licenses its GenAI system to a third party, the covered provider shall require both of the following as terms of the license: shall, when it provides the license, notify the licensee of the licensees obligations under this chapter.
(A)That the system remains in compliance with this chapter, to the extent it is technically feasible.(B)That the covered provider may revoke, suspend, or terminate the licensees authorization to use the GenAI system if the licensee modifies the GenAI system such that it no longer complies with this chapter.(2)If a covered provider knows that an identifiable third-party licensee modified a licensed GenAI system such that it no longer complies with this chapter, the covered provider shall terminate the licensees authorization to use the GenAI system within 72 hours of discovering the licensees action. do either of the following:
(A)Terminate the licensees authorization to use the GenAI system within 72 hours of discovering the licensees action.
(B)Notify the licensee of both of the following within 72 hours of discovering the licensees action:
(i)The licensees noncompliance.
(ii)The requirement to report noncompliance with paragraph (3) to the Attorney General pursuant to paragraph (4).
(3)A third-party licensee that receives a notice under paragraph (2) shall modify the licensed GenAI system to be in compliance with this chapter or cease using or making available a the licensed GenAI system, including a copy or modified version of the GenAI system, after the licensees authorization to use the GenAI system has been terminated by the covered provider pursuant to paragraph (2). system and shall report which of those actions it took to the covered provider within 96 hours of receipt of the notice.
(4)If a covered provider does not receive a report pursuant to paragraph (3) or if a covered provider receives a report from a third-party licensee indicating that the third-party licensee did not, pursuant to paragraph (3), modify or cease using or making available the licensed GenAI system, the covered provider shall report the third-party licensees noncompliance with paragraph (3) to the Attorney General.
(4)(5)This subdivision does not require a covered provider to monitor, investigate, or otherwise inquire into a third-party licensees use or modification of a licensed GenAI system.
(c)The Attorney General shall establish a mechanism to receive reports from covered providers submitted to the Attorney General pursuant to this section.
(d)A covered provider shall not falsely represent that a GenAI system is designed to primarily function as assistive technology.
SEC. 5.Section 22757.4 of the Business and Professions Code is amended to read:22757.4.(a)(1)A violator of this chapter shall be liable for a civil penalty in the amount of five thousand dollars ($5,000) per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel.
(2)A prevailing plaintiff in an action brought pursuant to this subdivision shall be entitled to all reasonable attorneys costs and fees.
(b)Each day that a covered provider, large online platform, or capture device manufacturer is in violation of this chapter shall be deemed a discrete violation.
(c)For a violation by a third-party licensee of paragraph (3) of subdivision (c) of Section 22757.3, the Attorney General, a county counsel, or a city attorney may bring a civil action for both of the following:(1)Injunctive relief.(2)Reasonable attorneys fees and costs.(c)A civil action brought under this section before the effective date of the act adding this subdivision shall not be maintained if the alleged conduct does not violate this chapter on and after the effective date of the act adding this subdivision.
(d)This section does not apply with respect to a violation of subdivision (d) of Section 22757.3.
SEC. 6.Section 22757.4.1 is added to the Business and Professions Code, to read:22757.4.1.(a)(1)A violator of subdivision (d) of Section 22757.3 shall be liable for a civil penalty in the amount of fifty thousand dollars ($50,000) per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel.
(2)A prevailing plaintiff in an action brought pursuant to this subdivision shall be entitled to all reasonable attorneys costs and fees.
(b)Each day that a covered provider is in violation of this section shall be deemed a discrete violation.
(c)This section shall remain in effect until January 1, 2029, and as of that date is repealed.
SEC. 7.Section 22757.5 of the Business and Professions Code is amended to read:22757.5.(a)This chapter does not apply to any product, service, internet website, or application that provides exclusively non-user-generated video game, television, streaming, movie, or interactive experiences. videogames.
(b)Before January 1, 2029, this chapter does not apply to a GenAI system that is designed to primarily function as assistive technology.
SEC. 5.SEC. 8.This act is an urgency statute necessary for the immediate preservation of the public peace, health, or safety within the meaning of Article IV of the California Constitution and shall go into immediate effect. The facts constituting the necessity are:
Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative artificial intelligence to create content for other malicious purposes, and accordingly, it is necessary that this act go into immediate effect.
The Legislature finds and declares all of the following:
(a)Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence (AI) system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative AI to create content for other malicious purposes.
(b)Conveying conflicting manifest and latent provenance information to consumers is detrimental to trust in both the provenance of content and the authenticity of any person, entity, or institution conveyed in that content.
(c)AI provenance technology is still developing, and ensuring that information conveyed to consumers regarding a piece of contents provenance is accurate, accessible, and resilient to common transformations to the extent technically feasible over time requires legislation to be adaptive and technologically agnostic.
(d)Accordingly, encouraging the adoption, interoperability, and continued development of widely accepted technical standards is necessary to promote reliable, scalable, and evolvable content provenance mechanisms.
SEC. 2.Section 22757.1 of the Business and Professions Code is amended to read:22757.1.As used in this chapter:
(a)Artificial intelligence or AI 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)Assistive technology means an item, piece of equipment, or product system, whether acquired commercially, modified, or customized, that is used to increase, maintain, or improve functional capabilities of individuals with disabilities and any service that directly assists an individual with a disability in the selection, acquisition, or use of the item, equipment, or product system.
(b)(c)Capture device means a device that can record photographs, audio, or video content, including, but not limited to, video and still photography cameras, mobile phones with built-in cameras or microphones, and voice recorders.
(c)(d)(1)Capture device manufacturer means a person who produces a capture device for sale in the state.
(2)Capture device manufacturer does not include a person exclusively engaged in the assembly of a capture device.
(d)(e)Covered provider means a person that creates, codes, or otherwise produces a generative artificial intelligence system that is publicly accessible within the geographic boundaries of the state.
(e)(f)Digital signature means a cryptography-based method that identifies the user or entity that attests to the information provided in the signed section.
(f)(g)Generative artificial intelligence system or GenAI system means an artificial intelligence that can generate derived synthetic content, including text, images, video, and audio, that emulates the structure and characteristics of the systems training data.
(g)(h)GenAI hosting platform means an internet website or application that makes available for download the source code or model weights a generative artificial intelligence system by a resident of the state, regardless of whether the terms of that use include compensation.
(h)(i)(1)Large online platform means a public-facing social media platform, file-sharing platform, mass messaging platform, or stand-alone search engine that distributes content to users who did not create or collaborate in creating the content that exceeded 2,000,000 unique monthly users during the preceding 12 months.
(2)Large online platform does not include either of the following:
(A)A broadband internet access service, as defined in Section 3100 of the Civil Code.
(B)A telecommunications service, as defined in Section 153 of Title 47 of the United States Code.
(i)Latent means present but not manifest.(j)Manifest means easily perceived, understood, or recognized by a natural person.(k)(j)Mass messaging platform means a direct messaging platform that allows users to distribute content to more than 100 users simultaneously.
(l)(k)Metadata means structural or descriptive information about data.
(m)(l)Minor modification means any of the following alterations:
(1)A change to brightness, contrast, or color.
(2)Sharpening.
(3)Saturating.
(4)Resizing. File resizing.
(5)Scaling.
(6)Cropping.
(7)Format File format conversions.
(8)Resampling.(9)(8)Denoising and removal of background noise in audio.
(n)(m)Personal information has the same meaning as defined in Section 1798.140 of the Civil Code.
(o)(n)Provenance data means information about the origin of a piece of digital content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
(p)(o)System provenance data means provenance data that is not reasonably capable of being associated with a particular user and that contains either of the following:
(1)Information regarding the type of device, system, or service that was used to generate a piece of digital content.
(2)Information related to content authenticity.
SEC. 3.Section 22757.2 of the Business and Professions Code is amended to read:22757.2.(a)A covered provider shall make available a disclosure verification tool at no cost to the user that meets all of the following criteria:
(1)The tool allows a user to assess whether image, video, or audio content, or content that is any combination thereof, was created or altered, except by minor modification, by the covered providers GenAI system.
(2)The tool outputs any system provenance data that is detected in the content.
(3)(A)Except as provided in subparagraph B, (B), the tool does not output any personal information that is detected in the content.
(B)The tool may output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content pursuant to a notice that does both of the following:
(i)Informs the user of the personal information that may be output by the tool.
(ii)Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the files permanent digital footprint and cannot be retracted from copies already in circulation.
(4)(A)Subject to subparagraph (B), the tool is publicly accessible.
(B)A covered provider may impose reasonable limitations on access to the tool to prevent, or respond to, demonstrable risks to the security or integrity of its GenAI system or to prevent misuse of the tool for malicious purposes.
(5)The tool allows a user to upload content or provide a uniform resource locator (URL) linking to online content.
(6)The tool supports technology, including an application programming interface, that allows a user to invoke the tool without visiting the covered providers internet website.
(b)A covered provider shall collect user feedback related to the efficacy of the covered providers AI disclosure verification tool and incorporate relevant feedback into any attempt to improve the efficacy of the tool.
(c)A (1)Except as provided in paragraph (2), a covered provider shall not collect, use, or retain retain, sell, share, or otherwise make available personal information derived either from a user of the covered providers AI disclosure verification tool or from any content submitted to processed by the disclosure verification tool beyond what is reasonably necessary for user authentication. strictly necessary to comply with this chapter.
(2)A covered provider may collect, retain, or use personal information for the sole purpose of communicating with a user that opts into being contacted by the covered provider.
(d)A covered provider shall not make access to the covered providers GenAI system or disclosure verification tool contingent upon a user providing personal information beyond what is strictly necessary for the purposes described in subparagraph (B) of paragraph (3) of subdivision (a). to comply with this chapter.
(e)A covered provider may satisfy the requirements of this section by directing a user to a third-party disclosure verification tool that is all of the following:
(1)In compliance with this section.
(2)Compatible with latent disclosures included in content produced or altered by the covered providers GenAI system.
(3)Clearly and conspicuously accessible through the user interface of the covered providers GenAI system.
SEC. 4.Section 22757.3 of the Business and Professions Code is amended to read:22757.3.(a)A To the extent it is technically feasible, a covered provider shall include a latent disclosure that meets all of the following criteria in AI-generated any image, video, or audio content, or content that is any combination thereof, created or altered, except by minor modification, by the covered providers GenAI system that meets all of the following criteria: system:
(1)To the extent that it is technically feasible and reasonable, the The disclosure conveys all of the following information, either directly or through a link to a permanent internet website:
(A)The name of the covered provider.
(B)The name and version number information of the GenAI system that created or altered, except by minor modification, altered the content.
(C)The time and date of the contents creation or alteration.
(D)A unique identifier.
(E)Whether the content is generated or modified by artificial intelligence. GenAI system created or altered the content.
(F)Whether the GenAI system is designed to primarily function as assistive technology.
(2)The disclosure is permanent or extraordinarily difficult to remove or tamper with, to the extent it is technically feasible. with.
(3)The disclosure is detectable by compatible with the covered providers AI detection disclosure verification tool.
(4)The disclosure is compliant or interoperable with widely accepted recognized industry standards.
(b)(1)If a covered provider licenses its GenAI system to a third party, the covered provider shall require by contract that the licensee ensures that the system includes provenance data that meets the criteria specified in subdivision (a), to the extent it is technically feasible. both of the following as terms of the license:
(A)That the system remains in compliance with this chapter, to the extent it is technically feasible.
(B)That the covered provider may revoke, suspend, or terminate the licensees authorization to use the GenAI system if the licensee modifies the GenAI system such that it no longer complies with this chapter.
(2)If a covered provider knows that a an identifiable third-party licensee modified a licensed GenAI system such that it is no longer capable of including a disclosure required by subdivision (a) in content the system creates or alters, complies with this chapter, the covered provider shall revoke terminate the license licensees authorization to use the GenAI system within 48 72 hours of discovering the licensees action.
(3)A third-party licensee shall cease using or making available a licensed GenAI system system, including a copy or modified version of the GenAI system, after the license for licensees authorization to use the GenAI system has been revoked terminated by the covered provider pursuant to paragraph (2).
(4)This subdivision does not require a covered provider to monitor, investigate, or otherwise inquire into a third-party licensees use or modification of a licensed GenAI system.
SEC. 5.Section 22757.5 of the Business and Professions Code is amended to read:22757.5.This chapter does not apply to a product, service, internet website, or application that provides exclusively nonuser-generated videogame systems incapable of producing highly realistic videos or images that a reasonable person could confuse with reality.SEC. 6.SEC. 5.This act is an urgency statute necessary for the immediate preservation of the public peace, health, or safety within the meaning of Article IV of the California Constitution and shall go into immediate effect. The facts constituting the necessity are:
Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative artificial intelligence to create content for other malicious purposes, and accordingly, it is necessary that this act go into immediate effect.
The Legislature finds and declares all of the following:
(a)Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence (AI) system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative AI to create content for other malicious purposes.
(b)Conveying conflicting manifest and latent provenance information to consumers is detrimental to trust in both the provenance of content and the authenticity of any person, entity, or institution conveyed in that content.
(c)AI provenance technology is still developing, and ensuring that information conveyed to consumers regarding a piece of contents provenance is accurate, accessible, and resilient to common transformations to the extent technically feasible over time requires legislation to be adaptive and technologically agnostic.
(d)Accordingly, encouraging the adoption, interoperability, and continued development of widely accepted technical standards is necessary to promote reliable, scalable, and evolvable content provenance mechanisms.
SEC. 2.Section 22757.1 of the Business and Professions Code is amended to read:22757.1.As used in this chapter:
(a)Artificial intelligence or AI 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)Capture device means a device that can record photographs, audio, or video content, including, but not limited to, video and still photography cameras, mobile phones with built-in cameras or microphones, and voice recorders.
(c)(1)Capture device manufacturer means a person who produces a capture device for sale in the state.
(2)Capture device manufacturer does not include a person exclusively engaged in the assembly of a capture device.
(d)Covered provider means a person that creates, codes, or otherwise produces a generative AI artificial intelligence system that is publicly accessible within the geographic boundaries of the state.
(e)Digital signature means a cryptography-based method that identifies the user or entity that attests to the information provided in the signed section.
(f)Generative artificial intelligence system or GenAI system means an artificial intelligence that can generate derived synthetic content, including text, images, video, and audio, that emulates the structure and characteristics of the systems training data.
(g)GenAI hosting platform means an internet website or application that makes available for download the source code or model weights a generative artificial intelligence system by a resident of the state, regardless of whether the terms of that use include compensation.
(h)(1)Large online platform means a public-facing social media platform, file-sharing platform, mass messaging platform, or stand-alone search engine that distributes content to users who did not create or collaborate in creating the content that exceeded 2,000,000 unique monthly users during the preceding 12 months.
(2)Large online platform does not include either of the following:
(A)A broadband internet access service, as defined in Section 3100 of the Civil Code.
(B)A telecommunications service, as defined in Section 153 of Title 47 of the United States Code.
(i)Latent means present but not manifest.
(j)Manifest means easily perceived, understood, or recognized by a natural person.
(k)Mass messaging platform means a direct messaging platform that allows users to distribute content to more than 100 users simultaneously.
(l)Metadata means structural or descriptive information about data.
(m)Minor modification means any of the following alterations:
(1)A change to brightness, contrast, or color.
(2)Sharpening.
(3)Saturating.
(4)Resizing.
(5)Scaling.
(6)Cropping.
(7)Format conversions.
(8)Resampling.
(9)Denoising and removal of background noise in audio.
(n)Personal information has the same meaning as defined in Section 1798.140 of the Civil Code.
(o)(1)Personal provenance data means provenance data that contains either of the following:(A)Personal information.(B)Unique device, system, or service information that is reasonably capable of being associated with a particular user.(2)Personal provenance data does not include information contained within a digital signature.(p)(o)Provenance data means information about the origin of a piece of content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
(q)(p)System provenance data means provenance data that is not reasonably capable of being associated with a particular user and that contains either of the following:
(1)Information regarding the type of device, system, or service that was used to generate a piece of digital content.
(2)Information related to content authenticity.
SEC. 3.Section 22757.2 of the Business and Professions Code is amended to read:22757.2.(a)A covered provider shall make available a disclosure verification tool at no cost to the user that meets all of the following criteria:
(1)The tool allows a user to assess whether image, video, or audio content, or content that is any combination thereof, was created or altered, except by minor modification, by the covered providers GenAI system.
(2)The tool outputs any system provenance data that is detected in the content.
(3)(A)Execpt Except as provided in subparagraph B, the tool does not output any personal information that is detected in the content.
(B)The tool may output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content pursuant to a notice that does both of the following:
(i)Informs the user of the personal information that may be output by the tool.
(ii)Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the files permanent digital footprint and cannot be retracted from copies already in circulation.
(4)(A)Subject to subparagraph (B), the tool is publicly accessible.
(B)A covered provider may impose reasonable limitations on access to the tool to prevent, or respond to, demonstrable risks to the security or integrity of its GenAI system or to prevent misuse of the tool for malicious purposes.
(5)The tool allows a user to upload content or provide a uniform resource locator (URL) linking to online content.
(6)The tool supports technology, including an application programming interface, that allows a user to invoke the tool without visiting the covered providers internet website.
(b)A covered provider shall collect user feedback related to the efficacy of the covered providers AI disclosure verification tool and incorporate relevant feedback into any attempt to improve the efficacy of the tool.
(c)A covered provider shall not collect, use, or retain personal information from a user of the covered providers AI disclosure verification tool or any content submitted to the disclosure verification tool beyond what is reasonably necessary for user authentication.
(d)A covered provider shall not make access to the covered providers GenAI system or disclosure verification tool contingent upon providing personal information beyond what is strictly necessary for the purposes described in subparagraph (B) of paragraph (3) of subdivision (a).
SEC. 4.Section 22757.3 of the Business and Professions Code is amended to read:22757.3.(a)A covered provider shall include a latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered providers GenAI system that meets all of the following criteria:
(1)To the extent that it is technically feasible and reasonable, the disclosure conveys all of the following information, either directly or through a link to a permanent internet website:
(A)The name of the covered provider.
(B)The name and version number of the GenAI system that created or altered, except by minor modification, the content.
(C)The time and date of the contents creation or alteration.
(D)A unique identifier.
(E)Whether the content is generated or modified by artificial intelligence.
(2)If the disclosure is not consistent with widely accepted industry standards, the disclosure is detectable by the covered providers disclosure verification tool.(3)(2)The disclosure is permanent or extraordinarily difficult to remove or tamper with, to the extent it is technically feasible.
(3)The disclosure is detectable by the covered providers AI detection tool.
(4)The disclosure is compliant or interoperable with widely accepted industry standards.
(b)(1)If a covered provider licenses its GenAI system to a third party, the covered provider shall require by contract that the licensee ensures that the system includes provenance data that meets the criteria specified in subdivision (a), to the extent it is technically feasible.
(2)If a covered provider knows that a third-party licensee modified a licensed GenAI system such that it is no longer capable of including a disclosure required by subdivision (a) in content the system creates or alters, the covered provider shall revoke the license within 48 hours of discovering the licensees action.
(3)A third-party licensee shall cease using a licensed GenAI system after the license for the system has been revoked by the covered provider pursuant to paragraph (2).
SEC. 5.Section 22757.5 of the Business and Professions Code is amended to read:22757.5.This chapter does not apply to either of the following: a product, service, internet website, or application that provides exclusively nonuser-generated videogame systems incapable of producing highly realistic videos or images that a reasonable person could confuse with reality.
(a)A product, service, internet website, or application that provides exclusively non-user-generated video game systems incapable of producing highly realistic videos or images.(b)A product, service, or application that is primarily marketed and intended to facilitate accessibility for disabled individuals and is designed to prevent content generated by the product, service, internet website, or application from being downloaded, stored, or otherwise appropriated, to the extent technically feasible.SEC. 6.This act is an urgency statute necessary for the immediate preservation of the public peace, health, or safety within the meaning of Article IV of the California Constitution and shall go into immediate effect. The facts constituting the necessity are:
Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative artificial intelligence to create content for other malicious purposes, and accordingly, it is necessary that this act go into immediate effect.
The Legislature finds and declares all of the following:
(a)Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence (AI) system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative AI to create content for other malicious purposes.
(b)Conveying conflicting manifest and latent provenance information to consumers is detrimental to trust in both the provenance of content and the authenticity of any person, entity, or institution conveyed in that content.
(c)AI provenance technology is still developing, and ensuring that information conveyed to consumers regarding a piece of contents provenance is accurate, accessible, and resilient to common transformations to the extent technically feasible over time requires legislation to be adaptive and technologically agnostic.
(d)Accordingly, encouraging the adoption, interoperability, and continued development of widely accepted technical standards is necessary to promote reliable, scalable, and evolvable content provenance mechanisms.
SEC. 2.Section 22757.1 of the Business and Professions Code is amended to read:22757.1.As used in this chapter:
(a)Artificial intelligence or AI 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)Capture device means a device that can record photographs, audio, or video content, including, but not limited to, video and still photography cameras, mobile phones with built-in cameras or microphones, and voice recorders.
(c)(1)Capture device manufacturer means a person who produces a capture device for sale in the state.
(2)Capture device manufacturer does not include a person exclusively engaged in the assembly of a capture device.
(d)Covered provider means a person that creates, codes, or otherwise produces a generative artificial intelligence AI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state.
(e)Digital signature means a cryptography-based method that identifies the user or entity that attests to the information provided in the signed section.
(f)Generative artificial intelligence system or GenAI system means an artificial intelligence that can generate derived synthetic content, including text, images, video, and audio, that emulates the structure and characteristics of the systems training data.
(g)GenAI hosting platform means an internet website or application that makes available for download the source code or model weights a generative artificial intelligence system by a resident of the state, regardless of whether the terms of that use include compensation.
(h)(1)Large online platform means a public-facing social media platform, file-sharing platform, mass messaging platform, or stand-alone search engine that distributes content to users who did not create or collaborate in creating the content that exceeded 2,000,000 unique monthly users during the preceding 12 months.
(2)Large online platform does not include either of the following:
(A)A broadband internet access service, as defined in Section 3100 of the Civil Code.
(B)A telecommunications service, as defined in Section 153 of Title 47 of the United States Code.
(i)Latent means present but not manifest.
(j)Manifest means easily perceived, understood, or recognized by a natural person.
(k)Mass messaging platform means a direct messaging platform that allows users to distribute content to more than 100 users simultaneously.
(l)Metadata means structural or descriptive information about data.
(m)Minor modification means any of the following alterations:
(1)A change to brightness, contrast, or color.
(2)Sharpening.
(3)Saturating.
(4)Resizing.
(5)Scaling.
(6)Cropping.
(7)Format conversions.
(8)Resampling.
(9)Denoising and removal of background noise in audio.
(m)(n)Personal information has the same meaning as defined in Section 1798.140 of the Civil Code.
(n)(o)(1)Personal provenance data means provenance data that contains either of the following:
(A)Personal information.
(B)Unique device, system, or service information that is reasonably capable of being associated with a particular user.
(2)Personal provenance data does not include information contained within a digital signature.
(o)(p)Provenance data means data that is embedded into digital content, or that is included in the digital contents metadata, for the purpose of verifying the digital contents authenticity, origin, or history of modification. information about the origin of a piece of content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
(p)(q)System provenance data means provenance data that is not reasonably capable of being associated with a particular user and that contains either of the following:
(1)Information regarding the type of device, system, or service that was used to generate a piece of digital content.
(2)Information related to content authenticity.
SEC. 3.Section 22757.2 of the Business and Professions Code is amended to read:22757.2.(a)A covered provider shall make available an AI detection a disclosure verification tool at no cost to the user that meets all of the following criteria:
(1)The tool allows a user to assess whether image, video, or audio content, or content that is any combination thereof, was created or altered altered, except by minor modification, by the covered providers GenAI system.
(2)The tool outputs any system provenance data that is detected in the content.
(3)The(3)(A)Execpt as provided in subparagraph B, the tool does not output any personal provenance data information that is detected in the content.
(B)The tool may output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information in the content pursuant to a notice that does both of the following:
(i)Informs the user of the personal information that may be output by the tool.
(ii)Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the files permanent digital footprint and cannot be retracted from copies already in circulation.
(4)(A)Subject to subparagraph (B), the tool is publicly accessible.
(B)A covered provider may impose reasonable limitations on access to the tool to prevent, or respond to, demonstrable risks to the security or integrity of its GenAI system. system or to prevent misuse of the tool for malicious purposes.
(5)The tool allows a user to upload content or provide a uniform resource locator (URL) linking to online content.
(6)The tool supports technology, including an application programming interface interface, that allows a user to invoke the tool without visiting the covered providers internet website.
(b)A covered provider shall collect user feedback related to the efficacy of the covered providers AI detection disclosure verification tool and incorporate relevant feedback into any attempt to improve the efficacy of the tool.
(c)A covered provider shall not do any of the following: collect, use, or retain personal information from a user of the covered providers AI disclosure verification tool or any content submitted to the disclosure verification tool beyond what is reasonably necessary for user authentication.
(1)(A)Except as provided in subparagraph (B), collect or retain personal information from users of the covered providers AI detection tool.(B)(i)A covered provider may collect and retain the contact information of a user who submits feedback pursuant to subdivision (b) if the user opts in to being contacted by the covered provider.(ii)User information collected pursuant to clause (i) shall be used only to evaluate and improve the efficacy of the covered providers AI detection tool.(2)Retain any content submitted to the AI detection tool for longer than is necessary to comply with this section.(3)Retain any personal provenance data from content submitted to the AI detection tool by a user.SEC. 4.Section 22757.3 of the Business and Professions Code is amended to read:22757.3.(a)A covered provider shall offer the user the option to include a manifest disclosure in image, video, or audio content, or content that is any combination thereof, created or altered by the covered providers GenAI system that meets all of the following criteria:(1)The disclosure identifies content as AI-generated content.(2)The disclosure is clear, conspicuous, appropriate for the medium of the content, and understandable to a reasonable person.(3)The disclosure is permanent or extraordinarily difficult to remove, to the extent it is technically feasible.(b)22757.3.(a)A covered provider shall include a latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered providers GenAI system that meets all of the following criteria:
(1)To the extent that it is technically feasible and reasonable, the disclosure conveys all of the following information, either directly or through a link to a permanent internet website:
(A)The name of the covered provider.
(B)The name and version number of the GenAI system that created or altered altered, except by minor modification, the content.
(C)The time and date of the contents creation or alteration.
(D)A unique identifier.
(E)Whether the content is generated or modified by artificial intelligence.
(2)The If the disclosure is not consistent with widely accepted industry standards, the disclosure is detectable by the covered providers AI detection disclosure verification tool.
(3)The disclosure is consistent with widely accepted industry standards.(4)(3)The disclosure is permanent or extraordinarily difficult to remove, remove or tamper with, to the extent it is technically feasible.
(c)(b)(1)If a covered provider licenses its GenAI system to a third party, the covered provider shall require by contract that the licensee maintain ensures that the systems capability to include a disclosure required by subdivision (b) in content the system creates or alters. system includes provenance data that meets the criteria specified in subdivision (a), to the extent it is technically feasible.
(2)If a covered provider knows that a third-party licensee modified a licensed GenAI system such that it is no longer capable of including a disclosure required by subdivision (b) (a) in content the system creates or alters, the covered provider shall revoke the license within 96 48 hours of discovering the licensees action.
(3)A third-party licensee shall cease using a licensed GenAI system after the license for the system has been revoked by the covered provider pursuant to paragraph (2).
SEC. 5.Section 22757.5 of the Business and Professions Code is amended to read:22757.5.This chapter does not apply to any product, service, internet website, or application that provides exclusively non-user-generated video game, television, streaming, movie, or interactive experiences. either of the following:
(a)A product, service, internet website, or application that provides exclusively non-user-generated video game systems incapable of producing highly realistic videos or images.
(b)A product, service, or application that is primarily marketed and intended to facilitate accessibility for disabled individuals and is designed to prevent content generated by the product, service, internet website, or application from being downloaded, stored, or otherwise appropriated, to the extent technically feasible.
SEC. 6.This act is an urgency statute necessary for the immediate preservation of the public peace, health, or safety within the meaning of Article IV of the California Constitution and shall go into immediate effect. The facts constituting the necessity are:
Establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a generative artificial intelligence system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative artificial intelligence to create content for other malicious purposes, and accordingly, it is necessary that this act go into immediate effect.
|
Date
|
Status
|
|---|---|
| Sep 30, 2026 | Chaptered by Secretary of State. Chapter 861, Statutes of 2026. |
| Sep 30, 2026 | Approved by the Governor. |
| Sep 2, 2026 | Enrolled and presented to the Governor at 3 p.m. |
| Aug 27, 2026 | Assembly amendments concurred in. (Ayes 40. Noes 0. Page 6021.) Ordered to engrossing and enrolling. |
| Aug 27, 2026 | Assembly amendments concurred in. (Ayes 39. Noes 0.) Ordered to engrossing and enrolling. |
| Aug 25, 2026 | In Senate. Concurrence in Assembly amendments pending. |
| Aug 25, 2026 | Read third time. Urgency clause adopted. Passed. (Ayes 72. Noes 0. Page 6517.) Ordered to the Senate. |
| Aug 25, 2026 | Read third time. Urgency clause adopted. Passed. Ordered to the Senate. |
| Aug 21, 2026 | Assembly Rule 69(b)(1) suspended. (Ayes 55. Noes 17. Page 6296.) |
| Aug 21, 2026 | Ordered to third reading. |
| Aug 21, 2026 | Read third time and amended. |
| Aug 21, 2026 | Assembly Rule 69(b)(1) suspended. |
| Jul 2, 2026 | Read second time. Ordered to third reading. |
| Jul 1, 2026 | From committee: Do pass. (Ayes 10. Noes 0.) (July 1). |
| Jun 17, 2026 | From committee: Do pass and re-refer to Com. on APPR. (Ayes 15. Noes 0.) (June 16). Re-referred to Com. on APPR. |
| Jun 9, 2026 | From committee with author's amendments. Read second time and amended. Re-referred to Com. on P. & C.P. |
| May 26, 2026 | Referred to Com. on P. & C.P. |
| May 20, 2026 | In Assembly. Read first time. Held at Desk. |
| May 19, 2026 | Read third time. Urgency clause adopted. Passed. (Ayes 33. Noes 1. Page 4345.) Ordered to the Assembly. |
| May 19, 2026 | Read third time. Urgency clause adopted. Passed. (Ayes 33. Noes 1.) Ordered to the Assembly. |
| May 14, 2026 | From committee: Do pass. (Ayes 5. Noes 0. Page 4260.) (May 14). |
| May 14, 2026 | Read second time. Ordered to third reading. |
| May 14, 2026 | From committee: Do pass. (Ayes 5. Noes 0.) (May 14). |
| May 8, 2026 | Set for hearing May 14. |
| Apr 27, 2026 | April 27 hearing: Placed on APPR. suspense file. |
| Apr 17, 2026 | Set for hearing April 27. |
| Apr 14, 2026 | From committee: Do pass and re-refer to Com. on APPR. (Ayes 8. Noes 1. Page 3843.) (April 13). Re-referred to Com. on APPR. |
| Apr 14, 2026 | From committee: Do pass and re-refer to Com. on APPR. (Ayes 8. Noes 1.) (April 13). Re-referred to Com. on APPR. |
| Apr 13, 2026 | Set for hearing April 13. |
| Mar 26, 2026 | From committee with author's amendments. Read second time and amended. Re-referred to Com. on P., D.T., & C.P. |
| Feb 18, 2026 | Referred to Com. on P., D.T., & C.P. |
| Feb 10, 2026 | From printer. May be acted upon on or after March 12. |
| Feb 9, 2026 | Introduced. Read first time. To Com. on RLS. for assignment. To print. |
|
Date
|
Description
|
Location
|
Votes
|
|
|---|---|---|---|---|
| Aug 27, 2026 | Unfinished Business SB1000 Becker Urgency Clause Concurrence | Senate Floor |
Passed |
|
| Aug 25, 2026 | SB 1000 Becker Third Reading Urgency By Wicks | Assembly Floor |
Passed |
|
| Jul 1, 2026 | Do pass. | Assembly Standing Committee on Appropriations |
Passed |
|
| Jun 16, 2026 | Do pass and be re-referred to the Committee on [Appropriations] | Assembly Standing Committee on Privacy and Consumer Protection |
Passed |
|
| May 19, 2026 | Senate 3rd Reading SB1000 Becker Urgency Clause | Senate Floor |
Passed |
|
| May 14, 2026 | Do pass | Senate Standing Committee on Appropriations |
Passed |
|
| Apr 27, 2026 | Placed on suspense file | Senate Standing Committee on Appropriations |
Passed |
|
| Apr 13, 2026 | Do pass, but first be re-referred to the Committee on [Appropriations] | Senate Standing Committee on Privacy, Digital Technologies, and Consumer Protection |
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
|
|---|---|---|
| California Initiative for Technology & Democracy | ** Support | |
| Center for Ai and Digital Policy (CAIDP) | ** Support | |
| ForensicVB LLC | ** Support | |
| Technet | ** Support | |
| The California Initiative for Technology and Democracy (CITED) | ** Support | |
| The Center for Ai and Digital Policy | ** Support | |
| Transparency Coalition.ai | ** Support |
SENATE RULES COMMITTEE Office of Senate Floor Analyses (916) 651-4171 | SB 1000 |
UNFINISHED BUSINESS
Bill No: SB 1000
Author: Becker (D)
Amended: 8/21/26
Vote: 27 - Urgency
SENATE PRIV., DIGITAL TECH. & CONS. PROT. COMMITTEE: 8-1, 4/13/26
AYES: Cabaldon, Gonzalez, McNerney, Ochoa Bogh, Padilla, Reyes, Umberg, Wiener
NOES: Jones
SENATE APPROPRIATIONS COMMITTEE: 5-0, 5/14/26
AYES: Cervantes, Cabaldon, Grayson, Richardson, Wahab
NO VOTE RECORDED: Seyarto, Dahle
SENATE FLOOR: 33-1, 5/19/26
AYES: Allen, Archuleta, Arreguín, Ashby, Becker, Blakespear, Cabaldon, Caballero, Cervantes, Cortese, Durazo, Gonzalez, Grayson, Grove, Hurtado, Laird, Limón, McGuire, McNerney, Menjivar, Ochoa Bogh, Padilla, Pérez, Reyes, Richardson, Rubio, Smallwood-Cuevas, Stern, Umberg, Valladares, Wahab, Weber Pierson, Wiener
NOES: Choi
NO VOTE RECORDED: Alvarado-Gil, Dahle, Jones, Niello, Seyarto, Strickland
ASSEMBLY FLOOR: 72-0, 8/25/26 - See last page for vote
SUBJECT: California AI Transparency Act
SOURCE: Author
DIGEST:
This bill makes a number of changes to the California AI Transparency (CAIT) Act, including eliminating the requirement to offer manifest disclosures, removing the user threshold on the definition of “covered provider,” amending the requirements applicable to third-party licensees, and amending other definitions and exemptions.
Assembly Amendments
refine privacy protections, specify that various disclosure verification tool requirements may be fulfilled through the use of third-party disclosure verification tools, as provided, require latent disclosures to specify whether a generative artificial intelligence system is designed to primarily function as assistive technology, prohibit making false representations, among other changes.
ANALYSIS:
Existing law:
- Establishes the CAIT Act, which becomes operative, in part, on August 2, 2026, and requires certain “covered providers” to make an artificial intelligence (AI) detection tool available at no cost by which a person can assess whether content was created or altered by the provider’s GenAI system. (Business (Bus.) & Professions (Prof.) Code § 22757 et seq.)
- Requires a “covered provider,” a person that creates, codes, or otherwise produces a GenAI system that has over 1 million monthly visitors or users and is publicly accessible within the geographic boundaries of the state, to offer users the option to include in AI-generated image, video, or audio content created by its own generative AI system a manifest disclosure that meets specified criteria, including that it identifies the content as AI-generated content. (Bus. & Prof. Code § 22757.3(a).)
- Requires a covered provider to include in AI-generated image, audio, and video content created by its generative AI system a latent disclosure that is detectable by the tool specified above and is, to the extent technically feasible, permanent or extraordinarily difficult to remove. (Bus. & Prof. Code § 22757.3(b).)
- Prohibits a covered provider from doing any of the following in carrying out the duties above:
- Collect or retain personal information when a person utilizes the covered provider’s AI detection tool, except that it may collect and retain the contact information of a person who submitted feedback.
- Retain any content submitted to the AI detection tool for longer than is necessary to comply with this law. (Bus. & Prof. Code § 22757.2(c).)
- Requires a large online platform, starting January 1, 2027, to do one of the following:
- Detect whether any provenance data that is compliant with widely adopted specifications adopted by an established standards-setting body is embedded into or attached to content distributed on the large online platform.
- Provide a user interface to disclose the availability of system provenance data that reliably indicates that the content was generated or substantially altered by a GenAI system or captured by a capture device. The user interface shall make clearly and conspicuously available to users information sufficient to identify the content’s authenticity, origin, or history of modification, including specified information such as whether provenance data is available.
- Allow a user to inspect all available system provenance data that is compliant with widely adopted specifications adopted by an established standards-setting body in an easily accessible manner by any of several specified means. (Bus. & Prof. Code § 22757.3.1.)
- Provides that violators of the above provisions are liable for a civil penalty in the amount of $5,000 per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel. Each day in violation is deemed a discrete violation. (Bus. & Prof. Code § 22757.4.)
- Provides that the CAIT Act does not apply to any product, service, website, or application that provides exclusively non-user-generated video game, television, streaming, movie, or interactive experiences.
This bill:
- Removes the user threshold from the definition of “covered provider.”
- Removes the requirement for covered providers to offer users the option to include a manifest disclosure in content. Amends the latent disclosure requirement only to the extent “technically feasible.”
- Redefines “provenance data” to mean information about the origin of a piece of digital content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
- Replaces references to “AI detection tool” with “disclosure verification tool.”
- Provides that the tool need not detect altered content if only a “minor modification,” which includes any of the following alterations:
- A change to brightness, contrast, or color.
- Sharpening.
- Saturating.
- File resizing.
- Scaling.
- Cropping.
- File format conversions.
- Denoising and removal of background noise in audio.
- Authorizes covered providers to impose reasonable limitations on access to the tool to prevent misuse of the tool for malicious purposes.
- Reworks the provisions regarding a users’ personal information. A covered provider shall not make access to the covered provider’s GenAI system or disclosure verification tool contingent upon providing personal information beyond what is strictly necessary for the specified purposes.
- Reworks third-party licensing provisions. Permits covered providers to notify a noncompliant licensee instead of immediately terminating their license. If licensee fails to cure violation and report back to covered provider, requires covered provider to notify Attorney General.
- Makes other technical and clarifying changes to the CAIT Act.
- Includes an urgency clause.
Background
Certain forms of media – audio recordings, video recordings, and still images – can be powerful evidence of the truth. While such media have always been susceptible to some degree of manipulation, fakes were relatively easy to detect. The rapid advancement of AI technology, specifically the wide-scale introduction of GenAI models, has made it drastically cheaper and easier to produce synthetic content created, audio, images, text, and video recordings that are not real, but that are so realistic that they are virtually impossible to distinguish from authentic content.
Last session, the Legislature passed the California AI Transparency Act (CAIT), which, beginning August 2, 2026, regulates provenance data disclosure in AI-generated content. “Covered providers” are required to embed latent provenance disclosures in content generated by their GenAI systems and to offer manifest disclosures. Providers are required to make available an AI detection tool at no cost to the user that meets certain criteria, including that it allows a user to assess whether certain content was created or altered by the covered provider’s system.
This bill is a product of many stakeholder conversations that seeks to clean up the CAIT Act and ensure it accomplishes its intended goal of providing useful transparency to Californians. The bill is author-sponsored. It is supported by Transparency Coalition.AI and ForensicVB LLC. No timely opposition was received.
Comment
According to the author:
As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren’t yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g., file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established by SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren’t interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. In light of the low implementation cost of current provenance standards relative to costs of model development, this bill removes the user threshold from the definition of "covered provider” to expand obligations to any AI system which is publicly available within the state, and aligns the act's requirements to reflect the state of content provenance standards and international provenance disclosure regimes. This bill is critical to ensuring California’s content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from.
FISCAL EFFECT: Appropriation: No Fiscal Com.: Yes Local: No
According to the Senate Appropriations Committee:
- Unknown costs to the state funded trial court system (Trial Court Trust Fund, General Fund) to adjudicate civil actions. By expanding who may be considered a covered provider and, consequently, subject to civil penalties with statutory damages, this bill may lead to additional case filings. The fiscal impact of this bill to the courts will depend on many unknowns, including the number of cases filed and the factors unique to each case. An eight-hour court day costs approximately $10,500 in staff workload. While the courts are not funded on a workload basis, an increase in workload could result in delayed court services and would put pressure on the General Fund to fund additional staff and resources and to increase the amount appropriated to backfill for trial court operations.
- The Department of Justice (DOJ) does not anticipate a significant fiscal impact.
- The California Department of Technology (CDT) notes that it would not fall within the definition of covered provider under this bill; therefore, it anticipates only minor and absorbable costs to continue tracking developments in the AI space.
According to the Assembly Appropriations Committee:
- Potential minor costs to the Department of Justice (General Fund) for enforcement. The Attorney General's authority to enforce the chapter already exists under current law, which this bill does not amend: a violation is subject to a civil penalty of $5,000 per violation, collected in a civil action filed by the Attorney General, a city attorney, or a county counsel. The bill does not create any new enforcement authority or mechanism. By removing the requirement that a covered provider have more than 1,000,000 monthly users, the bill enlarges the population of providers subject to the chapter, but any additional enforcement activity, and the associated costs, would likely be minor and would depend on whether the Attorney General elects to pursue actions under existing authority.
- The bill creates no new state program or appropriation, and the operative obligations apply to private covered providers. Enforcement by a city attorney or county counsel is permissive and would fall to local funds, so the bill imposes no reimbursable state-mandated local program. Any penalty revenue is not a significant or reliable funding source.
SUPPORT: (Verified 8/25/26)
Center for Ai and Digital Policy (CAIDP)
Forensicvb LLC
Transparency Coalition.ai
OPPOSITION: (Verified 8/25/26)
None received
ARGUMENTS IN SUPPORT: ForensicVB LLC writes:
From an operational standpoint, manifest disclosure or visible labels introduce a significant risk of investigative de-prioritization. When content is labeled—accurately or not—as AI-generated, it can create uncertainty at the front end of an investigation, delay triage decisions, and divert limited law enforcement resources away from time-sensitive CSAM cases. Put more simply, it has been reported that pedophiles operating on the Dark Web are labeling real CSAM content as “AI-generated” in the expectation that investigators will see the label and focus their attention on other non-labeled content.
For these reasons, I believe SB 1000’s current reliance on latent, provenance-based disclosure frameworks reflects a more responsible and technically sound path forward, and is consistent with existing content-provenance approaches already in use, such as the Coalition for Content Provenance and Authenticity (C2PA) specifications. The same professional experience that has led me to oppose manifest disclosure requirements has also led me to support latent disclosure approaches as a more reliable, durable, and effective trust signal for consumers of digital content. When implemented correctly, latent disclosure mechanisms can support authenticity, attribution, and accountability without creating avoidable investigative friction or opportunities for misuse.
ASSEMBLY FLOOR: 72-0 8/25/26
AYES: Addis, Aguiar-Curry, Ahrens, Alanis, Alvarez, Arambula, Ávila Farías, Bains, Bauer-Kahan, Bennett, Berman, Boerner, Bonta, Bryan, Calderon, Caloza, Carrillo, Chen, Connolly, Davies, DeMaio, Dixon, Ellis, Flora, Fong, Gabriel, Garcia, Gipson, Jeff Gonzalez, Mark González, Haney, Harabedian, Hart, Hoover, Irwin, Jackson, Johnson, Kalra, Krell, Lee, Lowenthal, Macedo, Muratsuchi, Nguyen, Ortega, Pacheco, Papan, Patel, Patterson, Pellerin, Petrie-Norris, Quirk-Silva, Ramos, Ransom, Celeste Rodriguez, Michelle Rodriguez, Rogers, Blanca Rubio, Sanchez, Schiavo, Schultz, Sharp-Collins, Solache, Soria, Stefani, Valencia, Wallis, Ward, Wicks, Wilson, Zbur, Rivas
NO VOTE RECORDED: Castillo, Elhawary, Hadwick, Lackey, McKinnor, Ta, Tangipa
Prepared by: Christian Kurpiewski / P., D.T., & C.P. / (916) 651-1548, Benjamin Diaz / P., D.T., & C.P. / (916) 651-1548
9/14/26 10:40:49
**** END ****
SENATE RULES COMMITTEE Office of Senate Floor Analyses (916) 651-4171 | SB 1000 |
UNFINISHED BUSINESS
Bill No: SB 1000
Author: Becker (D)
Amended: 8/21/26
Vote: 27 - Urgency
SENATE PRIV., DIGITAL TECH. & CONS. PROT. COMMITTEE: 8-1, 4/13/26
AYES: Cabaldon, Gonzalez, McNerney, Ochoa Bogh, Padilla, Reyes, Umberg, Wiener
NOES: Jones
SENATE APPROPRIATIONS COMMITTEE: 5-0, 5/14/26
AYES: Cervantes, Cabaldon, Grayson, Richardson, Wahab
NO VOTE RECORDED: Seyarto, Dahle
SENATE FLOOR: 33-1, 5/19/26
AYES: Allen, Archuleta, Arreguín, Ashby, Becker, Blakespear, Cabaldon, Caballero, Cervantes, Cortese, Durazo, Gonzalez, Grayson, Grove, Hurtado, Laird, Limón, McGuire, McNerney, Menjivar, Ochoa Bogh, Padilla, Pérez, Reyes, Richardson, Rubio, Smallwood-Cuevas, Stern, Umberg, Valladares, Wahab, Weber Pierson, Wiener
NOES: Choi
NO VOTE RECORDED: Alvarado-Gil, Dahle, Jones, Niello, Seyarto, Strickland
ASSEMBLY FLOOR: Passed, 8/25/26
SUBJECT: California AI Transparency Act
SOURCE: Author
DIGEST:
This bill makes a number of changes to the California AI Transparency (CAIT) Act, including eliminating the requirement to offer manifest disclosures, removing the user threshold on the definition of “covered provider,” amending the requirements applicable to third-party licensees, and amending other definitions and exemptions.
Assembly Amendments
refine privacy protections, specify that various disclosure verification tool requirements may be fulfilled through the use of third-party disclosure verification tools, as provided, require latent disclosures to specify whether a generative artificial intelligence system is designed to primarily function as assistive technology, prohibit making false representations, among other changes.
ANALYSIS:
Existing law:
- Establishes the CAIT Act, which becomes operative, in part, on August 2, 2026, and requires certain “covered providers” to make an artificial intelligence (AI) detection tool available at no cost by which a person can assess whether content was created or altered by the provider’s GenAI system. (Business (Bus.) & Professions (Prof.) Code § 22757 et seq.)
- Requires a “covered provider,” a person that creates, codes, or otherwise produces a GenAI system that has over 1 million monthly visitors or users and is publicly accessible within the geographic boundaries of the state, to offer users the option to include in AI-generated image, video, or audio content created by its own generative AI system a manifest disclosure that meets specified criteria, including that it identifies the content as AI-generated content. (Bus. & Prof. Code § 22757.3(a).)
- Requires a covered provider to include in AI-generated image, audio, and video content created by its generative AI system a latent disclosure that is detectable by the tool specified above and is, to the extent technically feasible, permanent or extraordinarily difficult to remove. (Bus. & Prof. Code § 22757.3(b).)
- Prohibits a covered provider from doing any of the following in carrying out the duties above:
- Collect or retain personal information when a person utilizes the covered provider’s AI detection tool, except that it may collect and retain the contact information of a person who submitted feedback.
- Retain any content submitted to the AI detection tool for longer than is necessary to comply with this law. (Bus. & Prof. Code § 22757.2(c).)
- Requires a large online platform, starting January 1, 2027, to do one of the following:
- Detect whether any provenance data that is compliant with widely adopted specifications adopted by an established standards-setting body is embedded into or attached to content distributed on the large online platform.
- Provide a user interface to disclose the availability of system provenance data that reliably indicates that the content was generated or substantially altered by a GenAI system or captured by a capture device. The user interface shall make clearly and conspicuously available to users information sufficient to identify the content’s authenticity, origin, or history of modification, including specified information such as whether provenance data is available.
- Allow a user to inspect all available system provenance data that is compliant with widely adopted specifications adopted by an established standards-setting body in an easily accessible manner by any of several specified means. (Bus. & Prof. Code § 22757.3.1.)
- Provides that violators of the above provisions are liable for a civil penalty in the amount of $5,000 per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel. Each day in violation is deemed a discrete violation. (Bus. & Prof. Code § 22757.4.)
- Provides that the CAIT Act does not apply to any product, service, website, or application that provides exclusively non-user-generated video game, television, streaming, movie, or interactive experiences.
This bill:
- Removes the user threshold from the definition of “covered provider.”
- Removes the requirement for covered providers to offer users the option to include a manifest disclosure in content. Amends the latent disclosure requirement only to the extent “technically feasible.”
- Redefines “provenance data” to mean information about the origin of a piece of digital content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
- Replaces references to “AI detection tool” with “disclosure verification tool.”
- Provides that the tool need not detect altered content if only a “minor modification,” which includes any of the following alterations:
- A change to brightness, contrast, or color.
- Sharpening.
- Saturating.
- File resizing.
- Scaling.
- Cropping.
- File format conversions.
- Denoising and removal of background noise in audio.
- Authorizes covered providers to impose reasonable limitations on access to the tool to prevent misuse of the tool for malicious purposes.
- Reworks the provisions regarding a users’ personal information. A covered provider shall not make access to the covered provider’s GenAI system or disclosure verification tool contingent upon providing personal information beyond what is strictly necessary for the specified purposes.
- Reworks third-party licensing provisions. Permits covered providers to notify a noncompliant licensee instead of immediately terminating their license. If licensee fails to cure violation and report back to covered provider, requires covered provider to notify Attorney General.
- Makes other technical and clarifying changes to the CAIT Act.
- Includes an urgency clause.
Background
Certain forms of media – audio recordings, video recordings, and still images – can be powerful evidence of the truth. While such media have always been susceptible to some degree of manipulation, fakes were relatively easy to detect. The rapid advancement of AI technology, specifically the wide-scale introduction of GenAI models, has made it drastically cheaper and easier to produce synthetic content created, audio, images, text, and video recordings that are not real, but that are so realistic that they are virtually impossible to distinguish from authentic content.
Last session, the Legislature passed the California AI Transparency Act (CAIT), which, beginning August 2, 2026, regulates provenance data disclosure in AI-generated content. “Covered providers” are required to embed latent provenance disclosures in content generated by their GenAI systems and to offer manifest disclosures. Providers are required to make available an AI detection tool at no cost to the user that meets certain criteria, including that it allows a user to assess whether certain content was created or altered by the covered provider’s system.
This bill is a product of many stakeholder conversations that seeks to clean up the CAIT Act and ensure it accomplishes its intended goal of providing useful transparency to Californians. The bill is author-sponsored. It is supported by Transparency Coalition.AI and ForensicVB LLC. No timely opposition was received.
Comment
According to the author:
As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren’t yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g., file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established by SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren’t interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. In light of the low implementation cost of current provenance standards relative to costs of model development, this bill removes the user threshold from the definition of "covered provider” to expand obligations to any AI system which is publicly available within the state, and aligns the act's requirements to reflect the state of content provenance standards and international provenance disclosure regimes. This bill is critical to ensuring California’s content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from.
FISCAL EFFECT: Appropriation: No Fiscal Com.: Yes Local: No
According to the Senate Appropriations Committee:
- Unknown costs to the state funded trial court system (Trial Court Trust Fund, General Fund) to adjudicate civil actions. By expanding who may be considered a covered provider and, consequently, subject to civil penalties with statutory damages, this bill may lead to additional case filings. The fiscal impact of this bill to the courts will depend on many unknowns, including the number of cases filed and the factors unique to each case. An eight-hour court day costs approximately $10,500 in staff workload. While the courts are not funded on a workload basis, an increase in workload could result in delayed court services and would put pressure on the General Fund to fund additional staff and resources and to increase the amount appropriated to backfill for trial court operations.
- The Department of Justice (DOJ) does not anticipate a significant fiscal impact.
- The California Department of Technology (CDT) notes that it would not fall within the definition of covered provider under this bill; therefore, it anticipates only minor and absorbable costs to continue tracking developments in the AI space.
According to the Assembly Appropriations Committee:
- Potential minor costs to the Department of Justice (General Fund) for enforcement. The Attorney General's authority to enforce the chapter already exists under current law, which this bill does not amend: a violation is subject to a civil penalty of $5,000 per violation, collected in a civil action filed by the Attorney General, a city attorney, or a county counsel. The bill does not create any new enforcement authority or mechanism. By removing the requirement that a covered provider have more than 1,000,000 monthly users, the bill enlarges the population of providers subject to the chapter, but any additional enforcement activity, and the associated costs, would likely be minor and would depend on whether the Attorney General elects to pursue actions under existing authority.
- The bill creates no new state program or appropriation, and the operative obligations apply to private covered providers. Enforcement by a city attorney or county counsel is permissive and would fall to local funds, so the bill imposes no reimbursable state-mandated local program. Any penalty revenue is not a significant or reliable funding source.
SUPPORT: (Verified 8/25/26)
Center for Ai and Digital Policy (CAIDP)
Forensicvb LLC
Transparency Coalition.ai
OPPOSITION: (Verified 8/25/26)
None received
ARGUMENTS IN SUPPORT: ForensicVB LLC writes:
From an operational standpoint, manifest disclosure or visible labels introduce a significant risk of investigative de-prioritization. When content is labeled—accurately or not—as AI-generated, it can create uncertainty at the front end of an investigation, delay triage decisions, and divert limited law enforcement resources away from time-sensitive CSAM cases. Put more simply, it has been reported that pedophiles operating on the Dark Web are labeling real CSAM content as “AI-generated” in the expectation that investigators will see the label and focus their attention on other non-labeled content.
For these reasons, I believe SB 1000’s current reliance on latent, provenance-based disclosure frameworks reflects a more responsible and technically sound path forward, and is consistent with existing content-provenance approaches already in use, such as the Coalition for Content Provenance and Authenticity (C2PA) specifications. The same professional experience that has led me to oppose manifest disclosure requirements has also led me to support latent disclosure approaches as a more reliable, durable, and effective trust signal for consumers of digital content. When implemented correctly, latent disclosure mechanisms can support authenticity, attribution, and accountability without creating avoidable investigative friction or opportunities for misuse.
Prepared by: Christian Kurpiewski / P., D.T., & C.P. / (916) 651-1548, Benjamin Diaz / P., D.T., & C.P. / (916) 651-1548
8/25/26 18:16:37
**** END ****
SENATE THIRD READING
STR Bill Id:SB 1000
Author:(Becker)
As Amended Ver:August 21, 2026
2/3 vote. Urgency
SUMMARY
Amends the AI Transparency Act to address stakeholder concerns, including by accounting for common digital licensing practices by establishing a "notice-and-cure" provision for noncompliant licensees.
Major Provisions
- Expands the Act to cover all GenAI systems, not only those that have over 1 million monthly visitors or users.
- Removes a requirement for GenAI providers to allow users to include manifest disclosures in outputs.
- Allows GenAI providers to direct users to third-party disclosure verification tools that meets specified requirements.
- Clarifies that the Act does not apply to "minor modifications" of content by GenAI systems.
- Clarifies that the Act's latent disclosure requirements apply only to the extent they are technically feasible.
- Requires latent disclosures to include whether a GenAI system is designed to primarily function as assistive technology, as defined.
- Updates third-party licensing language to ensure compatibility with common licensing practices. To account for open-source licensing practices, permits covered providers to notify a noncompliant licensee instead of immediately terminating their license. If licensee fails to cure violation and report back to covered provider, requires covered provider to notify Attorney General.
- Requires the Attorney General to establish a mechanism to receive reports from covered providers pursuant to (7).
- Makes various technical and non-substantive changes.
COMMENTS
Background. Over the past three years, as generative artificial intelligence (GenAI) technologies have become more realistic and accessible, online content that appears genuine, but is actually false, has flooded social media and other large online platforms. This unmitigated spread of synthetic content threatens to harm Californians in numerous ways, such as through the proliferation of nonconsensual deepfake pornography, scams, and the distribution of targeted political disinformation.
California has sought to tackle the issue of GenAI-produced online content through a three-step plan: label all GenAI content "fake," label a significant portion of real content “real,” and prominently display these labels online. Beginning this August, SB 942 (Becker, 2024) requires large GenAI platforms to include machine-detectable "latent disclosures" in their outputs, effectively labeling this content "fake."
What this bill would do. This bill would update the AI Transparency Act to reflect stakeholder feedback prior to that law going into effect on August 2nd.
Analysis. The development of GenAI is creating exciting opportunities to grow California’s economy and improve the lives of its residents. GenAI can generate compelling text, images and audio in an instant – but with novel technologies come novel safety concerns. Image manipulation and video doctoring have existed for nearly as long as photography and recording equipment, but they have historically required great effort and talent. In the past few years, the rapid development of GenAI has drastically reduced those barriers to entry, allowing a vast quantity of convincing, but ultimately fake, content to be generated in an instant. The creation of imagery, video, and audio by GenAI has the potential to change the world by automating repetitive tasks and fostering creativity. When employed by bad actors, however, these capabilities have the potential to destroy lives and destabilize societies.
Many of the issues associated with deepfakes could be resolved, if only there were a reliable way to identify GenAI content. While at present no single solution exists, there are ongoing efforts to embed information related to "content provenance," the verifiable history of a piece of content, into both GenAI products and the products of real-life recorders, such as digital cameras. Under this framework, the users of social media platforms would be able to rely on provenance data to identify trustworthy content.
Seeking to enact a comprehensive content provenance framework to combat the rapid online proliferation of AI-generated content, the Legislature passed SB 942 (Becker, Ch. 291, Stats. 2024). The resulting "AI Transparency Act" applies to "covered providers" – developers of publicly-accessible GenAI systems with over one million monthly users – requiring them to provide a free, publicly accessible tool that allows users to detect whether audio, image, or video content was generated or altered by their systems. The act permits visible disclosures and requires latent, machine-detectable provenance data to be embedded into content generated by these providers’ GenAI systems, effectively requiring GenAI developers to disclose that content generated by their systems is "fake" beginning in mid-2026.
SB 1000 would amend the AI Transparency Act in response to stakeholder concerns prior to that law taking effect in August.
According to the Author
As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren't yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g. file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren't interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. This bill is critical to ensuring California's content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from.
Arguments in Support
The Center for AI and Digital Policy writes in support of a previous version of this bill:
SB 1000 is an appropriate expansion of the rights laid out in California’ AI Transparency Act. Specifically, it removes the 1,000,000 user threshold from the definition of ‘covered provider’, expands the latent disclosure requirement to include whether content is generated or modified by AI, and replaces the ‘AI detection tool’ with a ‘disclosure verification tool’, shifting the obligation from passive detection to active and embedded disclosure. Taken together, these provisions establish meaningful protections for California consumers.
Arguments in Opposition
None on file.
FISCAL COMMENTS
According to the Assembly Appropriations Committee's analysis of this bill:
"Potential minor costs to the Department of Justice (General Fund) for enforcement."
VOTES
SENATE FLOOR: 33-1-6
YES: Allen, Archuleta, Arreguín, Ashby, Becker, Blakespear, Cabaldon, Caballero, Cervantes, Cortese, Durazo, Gonzalez, Grayson, Grove, Hurtado, Laird, Limón, McGuire, McNerney, Menjivar, Ochoa Bogh, Padilla, Pérez, Reyes, Richardson, Rubio, Smallwood-Cuevas, Stern, Umberg, Valladares, Wahab, Weber Pierson, Wiener
NO: Choi
ABS, ABST OR NV: Alvarado-Gil, Dahle, Jones, Niello, Seyarto, Strickland
ASM PRIVACY AND CONSUMER PROTECTION: 15-0-0
YES: Bauer-Kahan, Macedo, Bryan, DeMaio, Hoover, Irwin, Lowenthal, McKinnor, Ortega, Patterson, Pellerin, Petrie-Norris, Ward, Wicks, Wilson
ASM APPROPRIATIONS: 10-0-5
YES: Wicks, Hoover, Arambula, Caloza, Dixon, Fong, Mark González, Krell, Pacheco, Solache
ABS, ABST OR NV: Calderon, Pellerin, Sharp-Collins, Ta, Tangipa
UPDATED
VERSION: August 21, 2026
CONSULTANT: Slater Sharp / P. & C.P. / (916) 319-2200 FN: 0003907
SENATE THIRD READING
STR Bill Id:SB 1000
Author:(Becker)
As Amended Ver:June 9, 2026
2/3 vote. Urgency
SUMMARY
Amends the AI Transparency Act to address stakeholder concerns prior to that law going into effect on August 2.
Major Provisions
- Expands the Act to cover all GenAI systems, not only those that have over 1 million monthly visitors or users.
- Removes a requirement for GenAI providers to allow users to include manifest disclosures in outputs.
- Allows GenAI providers to direct users to third-party disclosure verification tools that meets specified requirements.
- Clarifies that the Act does not apply to "minor modifications" of content by GenAI systems.
- Clarifies that the Act's latent disclosure requirements apply only to the extent they are technically feasible.
- Requires latent disclosures to include whether a GenAI system is designed to primarily function as assistive technology, as defined.
- Updates third-party licensing language to ensure compatibility with common licensing practices
- Makes various technical and non-substantive changes.
COMMENTS
Background. Over the past three years, as generative artificial intelligence (GenAI) technologies have become more realistic and accessible, online content that appears genuine, but is actually false, has flooded social media and other large online platforms. This unmitigated spread of synthetic content threatens to harm Californians in numerous ways, such as through the proliferation of nonconsensual deepfake pornography, scams, and the distribution of targeted political disinformation. California has sought to tackle the issue of GenAI-produced online content through a three-step plan: label all GenAI content "fake," label a significant portion of real content "real," and prominently display these labels online. Beginning this August, SB 942 (Becker) Chapter 291, Statutes of 2024, requires large GenAI platforms to include machine-detectable "latent disclosures" in their outputs, effectively labeling this content "fake." What this bill would do. This bill would update the AI Transparency Act to reflect stakeholder feedback prior to that law going into effect on August 2nd. Analysis. The development of GenAI is creating exciting opportunities to grow California’s economy and improve the lives of its residents. GenAI can generate compelling text, images and audio in an instant – but with novel technologies come novel safety concerns. Image manipulation and video doctoring have existed for nearly as long as photography and recording equipment, but they have historically required great effort and talent. In the past few years, the rapid development of GenAI has drastically reduced those barriers to entry, allowing a vast quantity of convincing, but ultimately fake, content to be generated in an instant. The creation of imagery, video, and audio by GenAI has the potential to change the world by automating repetitive tasks and fostering creativity. When employed by bad actors, however, these capabilities have the potential to destroy lives and destabilize societies. Many of the issues associated with deepfakes could be resolved, if only there were a reliable way to identify GenAI content. While at present no single solution exists, there are ongoing efforts to embed information related to "content provenance," the verifiable history of a piece of content, into both GenAI products and the products of real-life recorders, such as digital cameras. Under this framework, the users of social media platforms would be able to rely on provenance data to identify trustworthy content. Seeking to enact a comprehensive content provenance framework to combat the rapid online proliferation of AI-generated content, the Legislature passed SB 942 (Becker), Chapter 291, Statutes of 2024. The resulting "AI Transparency Act" applies to "covered providers" – developers of publicly-accessible GenAI systems with over one million monthly users – requiring them to provide a free, publicly accessible tool that allows users to detect whether audio, image, or video content was generated or altered by their systems. The act permits visible disclosures and requires latent, machine-detectable provenance data to be embedded into content generated by these providers' GenAI systems, effectively requiring GenAI developers to disclose that content generated by their systems is "fake" beginning in mid-2026. SB 1000 would amend the AI Transparency Act in response to stakeholder concerns prior to that law taking effect in August.
According to the Author
"As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren't yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g. file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) [Chapter 674, Statutes of 2025] established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren't interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. This bill is critical to ensuring California's content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from."
Arguments in Support
The Center for AI and Digital Policy writes in support of a previous version of this bill:
SB 1000 is an appropriate expansion of the rights laid out in California' AI Transparency Act. Specifically, it removes the 1,000,000 user threshold from the definition of 'covered provider', expands the latent disclosure requirement to include whether content is generated or modified by AI, and replaces the 'AI detection tool' with a 'disclosure verification tool', shifting the obligation from passive detection to active and embedded disclosure. Taken together, these provisions establish meaningful protections for California consumers.
Arguments in Opposition
None on file.
FISCAL COMMENTS
According to the Assembly Appropriations Committee's analysis of this bill:
"Potential minor costs to the Department of Justice (General Fund) for enforcement."
VOTES
SENATE FLOOR: 33-1-6
YES: Allen, Archuleta, Arreguín, Ashby, Becker, Blakespear, Cabaldon, Caballero, Cervantes, Cortese, Durazo, Gonzalez, Grayson, Grove, Hurtado, Laird, Limón, McGuire, McNerney, Menjivar, Ochoa Bogh, Padilla, Pérez, Reyes, Richardson, Rubio, Smallwood-Cuevas, Stern, Umberg, Valladares, Wahab, Weber Pierson, Wiener
NO: Choi
ABS, ABST OR NV: Alvarado-Gil, Dahle, Jones, Niello, Seyarto, Strickland
ASM PRIVACY AND CONSUMER PROTECTION: 15-0-0
YES: Bauer-Kahan, Macedo, Bryan, DeMaio, Hoover, Irwin, Lowenthal, McKinnor, Ortega, Patterson, Pellerin, Petrie-Norris, Ward, Wicks, Wilson
ASM APPROPRIATIONS: 10-0-5
YES: Wicks, Hoover, Arambula, Caloza, Dixon, Fong, Mark González, Krell, Pacheco, Solache
ABS, ABST OR NV: Calderon, Pellerin, Sharp-Collins, Ta, Tangipa
UPDATED
VERSION: June 9, 2026
CONSULTANT: Slater Sharp / P. & C.P. / (916) 319-2200 FN: 0003281
SENATE THIRD READING
STR Bill Id:SB 1000
Author:(Becker)
As Amended Ver:June 9, 2026
2/3 vote. Urgency
SUMMARY
Amends the AI Transparency Act to address stakeholder concerns prior to that law going into effect on August 2.
Major Provisions
- Expands the Act to cover all GenAI systems, not only those that have over 1 million monthly visitors or users.
- Removes a requirement for GenAI providers to allow users to include manifest disclosures in outputs.
- Allows GenAI providers to direct users to third-party disclosure verification tools that meets specified requirements.
- Clarifies that the Act does not apply to "minor modifications" of content by GenAI systems.
- Clarifies that the Act's latent disclosure requirements apply only to the extent they are technically feasible.
- Requires latent disclosures to include whether a GenAI system is designed to primarily function as assistive technology, as defined.
- Updates third-party licensing language to ensure compatibility with common licensing practices
- Makes various technical and non-substantive changes.
COMMENTS
Background. Over the past three years, as generative artificial intelligence (GenAI) technologies have become more realistic and accessible, online content that appears genuine, but is actually false, has flooded social media and other large online platforms. This unmitigated spread of synthetic content threatens to harm Californians in numerous ways, such as through the proliferation of nonconsensual deepfake pornography, scams, and the distribution of targeted political disinformation. California has sought to tackle the issue of GenAI-produced online content through a three-step plan: label all GenAI content "fake," label a significant portion of real content "real," and prominently display these labels online. Beginning this August, SB 942 (Becker) Chapter 291, Statutes of 2024, requires large GenAI platforms to include machine-detectable "latent disclosures" in their outputs, effectively labeling this content "fake." What this bill would do. This bill would update the AI Transparency Act to reflect stakeholder feedback prior to that law going into effect on August 2nd. Analysis. The development of GenAI is creating exciting opportunities to grow California’s economy and improve the lives of its residents. GenAI can generate compelling text, images and audio in an instant – but with novel technologies come novel safety concerns. Image manipulation and video doctoring have existed for nearly as long as photography and recording equipment, but they have historically required great effort and talent. In the past few years, the rapid development of GenAI has drastically reduced those barriers to entry, allowing a vast quantity of convincing, but ultimately fake, content to be generated in an instant. The creation of imagery, video, and audio by GenAI has the potential to change the world by automating repetitive tasks and fostering creativity. When employed by bad actors, however, these capabilities have the potential to destroy lives and destabilize societies. Many of the issues associated with deepfakes could be resolved, if only there were a reliable way to identify GenAI content. While at present no single solution exists, there are ongoing efforts to embed information related to "content provenance," the verifiable history of a piece of content, into both GenAI products and the products of real-life recorders, such as digital cameras. Under this framework, the users of social media platforms would be able to rely on provenance data to identify trustworthy content. Seeking to enact a comprehensive content provenance framework to combat the rapid online proliferation of AI-generated content, the Legislature passed SB 942 (Becker), Chapter 291, Statutes of 2024. The resulting "AI Transparency Act" applies to "covered providers" – developers of publicly-accessible GenAI systems with over one million monthly users – requiring them to provide a free, publicly accessible tool that allows users to detect whether audio, image, or video content was generated or altered by their systems. The act permits visible disclosures and requires latent, machine-detectable provenance data to be embedded into content generated by these providers' GenAI systems, effectively requiring GenAI developers to disclose that content generated by their systems is "fake" beginning in mid-2026. SB 1000 would amend the AI Transparency Act in response to stakeholder concerns prior to that law taking effect in August.
According to the Author
"As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren't yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g. file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) [Chapter 674, Statutes of 2025] established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren't interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. This bill is critical to ensuring California's content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from."
Arguments in Support
The Center for AI and Digital Policy writes in support of a previous version of this bill:
SB 1000 is an appropriate expansion of the rights laid out in California' AI Transparency Act. Specifically, it removes the 1,000,000 user threshold from the definition of 'covered provider', expands the latent disclosure requirement to include whether content is generated or modified by AI, and replaces the 'AI detection tool' with a 'disclosure verification tool', shifting the obligation from passive detection to active and embedded disclosure. Taken together, these provisions establish meaningful protections for California consumers.
Arguments in Opposition
None on file.
FISCAL COMMENTS
According to the Assembly Appropriations Committee's analysis of this bill:
"Potential minor costs to the Department of Justice (General Fund) for enforcement."
VOTES
SENATE FLOOR: 33-1-6
YES: Allen, Archuleta, Arreguín, Ashby, Becker, Blakespear, Cabaldon, Caballero, Cervantes, Cortese, Durazo, Gonzalez, Grayson, Grove, Hurtado, Laird, Limón, McGuire, McNerney, Menjivar, Ochoa Bogh, Padilla, Pérez, Reyes, Richardson, Rubio, Smallwood-Cuevas, Stern, Umberg, Valladares, Wahab, Weber Pierson, Wiener
NO: Choi
ABS, ABST OR NV: Alvarado-Gil, Dahle, Jones, Niello, Seyarto, Strickland
ASM PRIVACY AND CONSUMER PROTECTION: 15-0-0
YES: Bauer-Kahan, Macedo, Bryan, DeMaio, Hoover, Irwin, Lowenthal, McKinnor, Ortega, Patterson, Pellerin, Petrie-Norris, Ward, Wicks, Wilson
ASM APPROPRIATIONS: 10-0-5
YES: Wicks, Hoover, Arambula, Caloza, Dixon, Fong, Mark González, Krell, Pacheco, Solache
ABS, ABST OR NV: Calderon, Pellerin, Sharp-Collins, Ta, Tangipa
UPDATED
VERSION: June 9, 2026
CONSULTANT: Slater Sharp / P. & C.P. / (916) 319-2200 FN: 0003281
Date of Hearing: July 1, 2026
ASSEMBLY COMMITTEE ON APPROPRIATIONS
Buffy Wicks, Chair
SBPCA Bill Id:SB 1000
Author:(Becker) – As Amended Ver:June 9, 2026
Policy Committee: | Privacy and Consumer Protection | Vote: | 15 - 0 |
|
| ||
|
|
Urgency: Yes State Mandated Local Program: No Reimbursable: No
SUMMARY:
This bill revises the California AI Transparency Act, which governs covered providers of generative artificial intelligence (GenAI) systems.
Specifically, this bill:
- Removes the requirement that a covered provider have more than 1,000,000 monthly visitors or users, so the chapter reaches any GenAI system that is publicly accessible in the state.
- Revises the required no-cost, consumer-facing tool — renamed a “disclosure verification tool” — that allows a user to assess whether content was created or altered by a covered provider's GenAI system, and adds and revises related privacy limits on the personal information a covered provider may collect or retain.
- Revises the latent disclosure requirements for AI-generated content and the contractual obligations a covered provider must impose on third-party licensees, including shortening the deadline to terminate a noncompliant licensee’s authorization from 96 to 72 hours.
- Adds an urgency clause.
FISCAL EFFECT:
- Potential minor costs to the Department of Justice (General Fund) for enforcement. The Attorney General's authority to enforce the chapter already exists under current law, which this bill does not amend: a violation is subject to a civil penalty of $5,000 per violation, collected in a civil action filed by the Attorney General, a city attorney, or a county counsel. The bill does not create any new enforcement authority or mechanism. By removing the requirement that a covered provider have more than 1,000,000 monthly users, the bill enlarges the population of providers subject to the chapter, but any additional enforcement activity, and the associated costs, would likely be minor and would depend on whether the Attorney General elects to pursue actions under existing authority.
- The bill creates no new state program or appropriation, and the operative obligations apply to private covered providers. Enforcement by a city attorney or county counsel is permissive and would fall to local funds, so the bill imposes no reimbursable state-mandated local program. Any penalty revenue is not a significant or reliable funding source.
COMMENTS:
- Purpose. According to the author:
As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren’t yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g. file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren’t interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. This bill is critical to ensuring California’s content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from.
- Background. The California AI Transparency Act, codified in the Business and Professions Code, requires a covered provider of a GenAI system to make available a free tool allowing users to assess whether content was generated or altered by the provider’s system, to offer disclosures identifying AI-generated content, and to impose disclosure-related obligations on third parties licensing the provider’s system. This bill refines those requirements. The most consequential change for the scope of the chapter is the removal of the 1,000,000-monthly-user threshold from the definition of “covered provider,” which extends coverage from the largest systems to any GenAI system publicly accessible in the state. The bill also reworks the detection tool into a “disclosure verification tool” with revised functionality and privacy limits, revises the information a latent disclosure must convey and the standards with which a disclosure must be compatible, and shortens the window — from 96 to 72 hours — within which a covered provider must terminate a licensee’s authorization after learning the licensee has modified a system out of compliance. Many of the new obligations are qualified as applying only “to the extent technically feasible.”
Analysis Prepared by: Shiran Zohar / APPR. / (916) 319-2081
Date of Hearing: June 16, 2026
ASSEMBLY COMMITTEE ON PRIVACY AND CONSUMER PROTECTION
Rebecca Bauer-Kahan, Chair
SBPCA Bill Id:SB 1000
Author:(Becker) – As Amended Ver:June 9, 2026
SENATE VOTE: 33-1
SUBJECT: California AI Transparency Act
SYNOPSIS
Over the past three years, as generative artificial intelligence (GenAI) technologies have become more realistic and accessible, online content that appears genuine, but is actually false, has flooded social media and other large online platforms. This unmitigated spread of synthetic content threatens to harm Californians in numerous ways, such as through the proliferation of nonconsensual deepfake pornography, scams, and the distribution of targeted political disinformation.
California has sought to tackle the issue of GenAI-produced online content through a three-step plan: label all GenAI content “fake,” label a significant portion of real content “real,” and prominently display these labels online. Beginning this August, SB 942 (Becker, 2024) requires large GenAI platforms to include machine-detectable “latent disclosures” in their outputs. This bill would update SB 942 to reflect stakeholder feedback prior to that law going into effect on August 2nd.
This bill is supported by Transparency Coalition.ai, Center for AI and Digital Policy, and ForensicVB LLC. California Initiative for Technology & Democracy, TechNet, and Disability Rights California take “support if amended” positions. The bill has no opposition.
EXISTING LAW:
- Defines the following terms:
- “Artificial intelligence” or “AI” 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. (Bus. & Prof. Code § 22757.1. et seq.)
- “Assistive technology” to mean an item, piece of equipment, or product system, whether acquired commercially, modified, or customized, that is used to increase, maintain, or improve functional capabilities of individuals with disabilities and any service that directly assists an individual with a disability in the selection, acquisition, or use of the item, equipment, or product system. (Welf. & Inst. Code § 19461.)
- “Covered provider” to mean a person that creates, codes, or otherwise produces a generative artificial intelligence system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state. (Bus. & Prof. Code § 22757.1. et seq.)
- “Generative artificial intelligence system” or “GenAI system” to mean an artificial intelligence that can generate derived synthetic content, including text, images, video, and audio, that emulates the structure and characteristics of the system’s training data. (Id.)
- “Latent” to mean present but not manifest. (Id.)
- “Manifest” to mean easily perceived, understood, or recognized by a natural person. (Id.)
- “Personal provenance data” to mean provenance data that contains either of the following:
- Personal information.
- Unique device, system, or service information that is reasonably capable of being associated with a particular user. (Bus. & Prof. Code § 22757.1. et seq.)
- Personal information.
- “Personal provenance data” does not include information contained within a digital signature. (Id.)
- “Provenance data” to mean data that is embedded into digital content, or that is included in the digital content’s metadata, for the purpose of verifying the digital content’s authenticity, origin, or history of modification. (Id.)
- “System provenance data” to mean provenance data that is not reasonably capable of being associated with a particular user and that contains either information regarding the type of device, system, or service that was used to generate a piece of digital content, or information related to content authenticity. (Id.)
- “Artificial intelligence” or “AI” 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. (Bus. & Prof. Code § 22757.1. et seq.)
- Beginning August 2, 2026, requires a covered provider to create and make freely available an AI detection tool that can be used to assess whether digital content was created or altered by the covered provider’s GenAI system, and that has various specified characteristics.
- Prohibits a covered provider from collecting or retaining personal information from users of the covered provider’s AI detection tool, or retaining personal provenance data from user-uploaded content, except as specified. (Bus. & Prof. Code § 22757.1. et seq.)
- Beginning August 2, 2026, requires a covered provider to offer a user of a GenAI tool the option to include a manifest disclosure in generated content, with various specified characteristics. (Id.)
- Beginning August 2, 2026, requires a covered provider to include a latent disclosure in AI-generated image, video, or audio content, or content that is any combination thereof, created by the covered provider’s GenAI system that, to the extent technically feasible and reasonable, conveys the name of the covered provider, the name and version number of the GenAI system that created or altered the content, the time and date of the content’s creation or alteration, and a unique identifier, either directly or through a link to a permanent internet website. (Id.)
- Requires that if a covered provider licenses their GenAI system to a third party, the covered provider shall require by contract that the licensee maintain the system’s capability to include a latent disclosure.
- Requires that if a covered provider knows a licensee has removed the capability of a licensed GenAI system to include a latent disclosure, the covered provider shall revoke the licensee’s contract.
- Requires that a licensee cease using a licensed GenAI system after its license has been revoked. (Id.)
- Exempts from the AI Transparency Act any product, service, internet website, or application that provides exclusively non-user-generated video game, television, streaming, movie, or interactive experience. (Id.)
THIS BILL:
- Amends the definitions in Section 22757.1 of the Business and Professions Code as follows:
- Adds “assistive technology,” defined to have the same meaning as in Welf. & Inst. Code § 19461.
- Adds “minor modification,” defined to mean any of the following alterations:
- A change to brightness, contrast, or color.
- Sharpening.
- Saturating.
- File resizing.
- Scaling.
- Cropping.
- File format conversions.
- Denoising and removal of background noise in audio.
- A change to brightness, contrast, or color.
- Removes “latent,” “manifest,” and “personal provenance data.”
- Updates “provenance data” to mean information about the origin of a piece of digital content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
- Makes various minor adjustments to the AI Transparency Act ahead of its August 2, 2026, implementation, including all of the following:
- Recasts “AI detection tool” as “disclosure verification tool.”
- Expands the act to cover all GenAI systems, not only those that have over 1 million monthly visitors or users.
- Clarifies that a disclosure verification tool shall allow users to assess whether content is created or altered, except by minor modification, by a covered provider’s GenAI system.
- Permits a disclosure verification tool to output personal information only if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content.
- Prohibits a covered provider from collecting, using, retaining, selling, sharing, or otherwise making available personal information derived either from a user of the covered provider’s disclosure verification tool or from any content processed by the disclosure verification tool beyond what is strictly necessary to comply with the chapter.
- Prohibits a covered provider from making access to the covered provider’s GenAI system or disclosure verification tool contingent upon a user providing personal information beyond what is strictly necessary to comply with the chapter.
- Allows a covered provider to comply with the requirement to make available a disclosure verification tool by instead directing a user to a third-party tool that complies with the chapter and is compatible with latent disclosures included in the covered provider’s GenAI content.
- Deletes the requirement that a covered provider offer a user the option to include a manifest disclosure in GenAI content.
- Clarifies that latent disclosures included in GenAI content pursuant to Section 22757.3 are only required to include specified information to the extent “technically feasible.”
- Recasts “version number” as “version information.”
- Adds a requirement for a latent disclosure to include whether or not a GenAI system is “designed to primarily function as assistive technology.”
- Recasts various third-party contracting requirements as “terms of [a] license” that are required to be included by a covered provider. Clarifies that covered providers are only required to respond to violations by an “identifiable” third party licensee and reduces the timeline for responding to 72 hours. Clarifies that a covered provider is not required to “monitor, investigate, or otherwise inquire into a third-party licensee’s use or modification of a licensed GenAI system” pursuant to the chapter.
- Includes an urgency clause.
- Makes various findings and declarations.
COMMENTS:
- Author’s statement. According to the author:
As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren’t yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g. file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren’t interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. This bill is critical to ensuring California’s content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from.
- Background. AI and GenAI. The development of GenAI is creating exciting opportunities to grow California’s economy and improve the lives of its residents. GenAI can generate compelling text, images and audio in an instant – but with novel technologies come novel safety concerns.
In brief, AI is the mimicking of human intelligence by artificial systems such as computers. AI uses algorithms – sets of rules – to transform inputs into outputs. Inputs and outputs can be anything a computer can process: numbers, text, audio, video, or movement. AI is not fundamentally different from other computer functions; its novelty lies in its application. Unlike traditional computer functions, AI can accomplish tasks that are normally performed by humans. AI that is trained on small, specific datasets to make recommendations and predictions is sometimes referred to as “predictive AI.” This differentiates it from GenAI, which is trained on massive datasets in order to produce detailed text and images. When Netflix suggests a TV show to a viewer, that recommendation is produced by predictive AI that has been trained on the viewing habits of Netflix users. When ChatGPT generates text in clear, concise paragraphs, it uses GenAI that has been trained on the written contents of the internet.
Deepfakes and Disinformation. Image manipulation and video doctoring have existed for nearly as long as photography and recording equipment, but they have historically required great effort and talent. In the past few years, the rapid development of GenAI has drastically reduced those barriers to entry, allowing a vast quantity of convincing, but ultimately fake, content to be generated in an instant. The creation of imagery, video, and audio by GenAI has the potential to change the world by automating repetitive tasks and fostering creativity. When employed by bad actors, however, these capabilities have the potential to destroy lives and destabilize societies.
Deepfake pornography. Since its inception, GenAI has been used to subject women and girls to image-based sexual abuse. While high-profile celebrities were most often targeted when this technology was first developed,[1] open-source GenAI models have been exploited to make this technology more accessible and affordable. This has led to a proliferation of websites and phone-based apps that offer user-friendly interfaces for uploading clothed images of real people to generate photorealistic nude images of not only adults, but also children. According to a New York Times article:
Boys in several states have used widely available “nudification” apps to pervert real, identifiable photos of their clothed female classmates, shown attending events like school proms, into graphic, convincing-looking images of the girls with exposed A.I.-generated breasts and genitalia. In some cases, boys shared the faked images in the school lunchroom, on the school bus or through group chats on platforms like Snapchat and Instagram, according to school and police reports.[2]
In February 2024, deepfake nude images of 16 eighth-grade students were circulated among students at a California middle school.[3] Similar reports of abuse have been reported across the country and show no sign of abating.[4] In the first six months of 2024, nudification sites had been visited over 200 million times.[5] Meanwhile, a 2024 study from Center on Democracy and Technology reports that 40% of students were aware of deepfakes being shared at school, 15% of which depicted an individual in a sexually explicit or intimate manner. In over 60% of these cases, the images were distributed via social media.[6]
Scams. GenAI-powered speech and video is driving a new era in scamming. These AI tools are often trained on publicly available data, and as a result, the more data a target has online, the easier it is to develop a passable imitation of them or their loved ones. This is especially true of wealthy clients, whose public appearances, including speeches, are often widely available on the internet.[7] For example, a complicated scam utilizing both deepfake video and false audio was recently performed in Hong Kong. A multinational company lost $25.6 million after employees were fooled by deepfake technology, with one incident involving a digitally recreated version of its chief financial officer ordering money transfers in a video conference call. Everyone present on the video call, except the victim, was a fake representation of real people. The scammers applied deepfake technology to turn publicly available video and other footage into convincing versions of the meeting’s participants.[8]
In December 2024, the FBI issued a public service announcement warning about the potential dangers posed by GenAI.[9] Among the concerns highlighted was the technology’s ability to significantly lower the barrier for producing counterfeit documents, such as fake IDs, passports, and other fraudulent government-issued identification, which could greatly facilitate identity theft. Additionally, GenAI enables the creation of entirely fictitious profiles on social media and dating platforms, which can be used to exploit individuals both financially and emotionally.
Elections. Deepfake technology is being used around the world to spread disinformation and propaganda. This has already been observed in Slovakia, where deepfake audio influenced an election in 2023:
Days before a pivotal election in Slovakia to determine who would lead the country, a damning audio recording spread online in which one of the top candidates seemingly boasted about how he’d rigged the election. And if that wasn’t bad enough, his voice could be heard on another recording talking about raising the cost of beer. The recordings immediately went viral on social media, and the candidate, who is pro-NATO and aligned with Western interests, was defeated in September by an opponent who supported closer ties to Moscow and Russian President Vladimir Putin.[10]
Similar deepfakes surfaced in the United States ahead of the 2024 presidential election. In July 2024, Elon Musk shared a video featuring an AI-generated voice clone of then-Vice President Kamala Harris, in which the fabricated voice claimed she was a “diversity hire” due to being a woman of color and that she “did not know the first thing about running a country.”[11] Although Musk admitted two days after posting that the video was intended as satire, the potential impact of such content on political campaigns remains a serious concern. In 2023, the Legislature enacted a trio of bills aimed at addressing elections deepfakes: AB 2655 (Berman, Ch. 261, Stats. 2024), which imposes removal and disclosure obligations on large online platforms during the four months leading up to an election; AB 2839 (Pellerin, Ch. 262, Stats. 2024), which prohibits the distribution of election materials containing certain types of deceptively altered digital content; and AB 2355 (Carillo, Ch. 260, Stats. 2024), which requires AI-altered campaign materials to include clear disclosures. The first two bills, however, have been blocked from being enacted by legal proceedings.
Content Provenance. Many of the issues associated with deepfakes could be resolved, if only there were a reliable way to identify GenAI content. While at present no single solution exists, there are ongoing efforts to embed information related to “content provenance,” the verifiable history of a piece of content, into both GenAI products and the products of real-life recorders, such as digital cameras. Under this framework, the users of social media platforms would be able to rely on provenance data to identify trustworthy content.
Metadata. The Merriam-Webster Dictionary defines metadata as “data that provides information about other data.” In practice, metadata is a structured set of descriptors attached to digital content. A photograph’s metadata might include information about the camera used to take the photo, the time and date the photo was taken, and the photo’s precise geolocation. A written document’s metadata may include details about its author, its creation date, and the number of times the document has been edited. Metadata can be used to verify content authenticity, helping to combat fake news by providing a traceable history of content creation and modification. But metadata can also contain personal information about the individual who creates or modifies a piece of content.
Watermarking. Watermarking is the process of embedding an identifiable marker into digital content. This “watermark” can be visible, such as a logo or text overlay, or it can be invisible, data embedded into a file in a manner that does not noticeably alter the content. As a technology, watermarking is in its infancy. Watermarks can be stripped from content relatively easily by common screenshotting tools, file compression software, and image editing programs like Photoshop. They can also be faked by treating a GenAI system like a copy machine: a real image or video can be fed into a GenAI system and spit back out, unaltered except for the addition of a watermark. The system’s user receives a piece of authentic content that has been incorrectly marked as inauthentic, ready to be posted online to create confusion and sow discord. Furthermore, while progress has been made towards developing standards for watermarking images and video, how text should be watermarked is far less clear.
SB 942. Seeking to enact a comprehensive content provenance framework to combat the rapid online proliferation of AI-generated content, the Legislature passed SB 942 (Becker, Ch. 291, Stats. 2024). The resulting “AI Transparency Act” applies to “covered providers” – developers of publicly-accessible GenAI systems with over one million monthly users – requiring them to provide a free, publicly accessible tool that allows users to detect whether audio, image, or video content was generated or altered by their systems. The act permits visible disclosures and requires latent, machine-detectable provenance data to be embedded into content generated by these providers’ GenAI systems, effectively requiring GenAI developers to disclose that content generated by their systems is “fake” beginning in mid-2026.
- What this bill would do. SB 1000 would amend the AI Transparency Act in response to stakeholder concerns. Major changes include:
- Expanding the bill to cover all GenAI systems, not only those that have over 1 million monthly visitors or users.
- Removing the requirement for GenAI providers to allow users to include manifest disclosures in outputs.
- Allowing GenAI providers to direct users to a third-party disclosure verification tool that meets specified requirements.
- Clarifying the bill does not apply to “minor modifications” of content by GenAI systems, as defined.
- Clarifying the bill’s latent disclosure requirements apply only to the extent they are technically feasible.
- Updating third-party licensing language to ensure compatibility with common licensing practices, specifying that GenAI providers only have to act on violations by identifiable licensees, reducing the response window from 96 hours to 72 hours, and clarifying that the bill’s third-party licensing provisions do not require GenAI providers to monitor third party use.
Writing in support, TransparencyCoalition.ai explains the need for this measure:
As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant societal and democratic risks, exacerbating problems such as disinformation, harassment, and fraud while threatening academic and journalistic integrity. At the same time, content provenance and disclosure technology is still in development, and researchers across both academia and industry are working to ensure methods of embedding information in content are scalable and capable of withstanding adversarial exploitation. It is therefore critical that provenance disclosure methods provide clear information to users to allow them to discern the trustworthiness and probable origin of content.
The bill contains an urgency clause allowing the AI Transparency Act to be updated immediately if the bill passes and the Governor signs it.
Accessibility Concerns. Disability Rights California has raised concerns about unintended consequences of the bill for individuals with disabilities. According to a June 9th letter to this Committee:
[The] bill as currently drafted does not adequately account for situations in which generative AI is used as an assistive technology by individuals with disabilities, particularly those with disabilities that impact their ability to communicate verbally . . . For individuals with conditions such as ALS, cerebral palsy, or other communication-related disabilities, AI-driven voice tools function as a substitute for natural speech. These tools are used in everyday contexts, including communicating with service providers, participating in professional or public discourse, and engaging in personal interactions. Applying AI-generated content labeling requirements to these communications risks mischaracterizing human-authored speech as synthetic content and may inadvertently disclose a person’s disability status.
Disability Rights California goes on to suggest exempting audio-based assistive technologies from the bill. Concerned that an outright exemption could enable bad actors to avoid disclosure requirements, the author instead recently amended SB 1000 to require that latent disclosures specify “whether [a] GenAI system is designed to primarily function as assistive technology.” Going forward, the author may wish to delay implementation of this requirement to give existing latent disclosure technologies time to adjust accordingly.
Video games. The AI Transparency Act includes the following exemption, meant to address concerns raised by the video game industry in response to SB 942:
“This chapter does not apply to any product, service, internet website, or application that provides exclusively non-user-generated video game, television, streaming, movie, or interactive experiences.”
An earlier version of SB 1000 would have modified the language of this exemption to more narrowly focus on video games incapable of outputting highly realistic content. Following further negotiations with industry and this committee, the author recently amended this bill such that it no longer changes the scope of this exemption.
ARGUMENTS IN SUPPORT:
Writing in support, the Center for AI and Digital Policy explains the need for this measure:
SB 1000 is an appropriate expansion of the rights laid out in California’ AI Transparency
Act. Specifically, it removes the 1,000,000 user threshold from the definition of ‘covered
provider’, expands the latent disclosure requirement to include whether content is generated or modified by AI, and replaces the ‘AI detection tool’ with a ‘disclosure verification tool’, shifting the obligation from passive detection to active and embedded disclosure. Taken together, these provisions establish meaningful protections for California consumers.
TechNet takes a “support if amended” position, writing the following:
The recent amendments to SB 1000 represent a notable improvement, and we value the author's commitment to refining the legislation. Given the necessity of addressing outstanding concerns with SB 942 prior to its August 2, 2026, activation, these updates are vital. The present version successfully harmonizes consumer safety, security, and technological progress by prioritizing latent disclosures over unfeasible manifest labeling requirements.
However, we maintain specific concerns regarding the bill as amended on June 9. Our feedback is intended to refine the legislation to ensure scalable compliance and avoid
unintended consequences, rather than to oppose its primary goals. We recommend additional clarity on the following points:
- We would like to see a clear exemption for business-to-business uses, because without it, covered providers that distribute or make available a generative artificial intelligence system to businesses or for business purposes face potential liability for content generated by downstream customers using their platforms—even though the provider has no knowledge of, control over, or intent regarding the specific content produced…
- Matching data minimization rules with practical tool functionality: The bill presently restricts the use of limited operational data needed to maintain and improve disclosure tools, relying too heavily on user feedback…
- Establishing direct compliance for third-party licensees: While we support uniform transparency protections, the bill currently places the burden of policing downstream compliance on model developers…
The California Initiative for Technology & Democracy takes a “support if amended” position, writing the following:
SB 1000 makes several improvements to the AI Transparency Act. First, it expands the coverage of the law by ensuring that all creators of generative AI systems are responsible for maintaining a functioning system to embed latent disclosures, underscoring the foundational importance of attaching content provenance to all AI-generated content. Second, the bill clarifies the exceptions to the bill to eliminate vague terms, including the undefined “interactive experiences.”
[…]
While CITED applauds the improvements outlined above, we believe there are still continued refinements that deserve attention. For instance, the bill excludes the listing of “minor modifications” in provenance information, a term defined by a listing of editing techniques.
[…]
Additionally, CITED objects to the elimination of the requirement for Gen AI Providers to provide users with the option to place manifest disclosures on the bill. Manifest disclosures remain the most accessible way a consumer can quickly discern the origins of digital content.
REGISTERED SUPPORT / OPPOSITION:
Support
Center for Ai and Digital Policy (CAIDP)
ForensicVB LLC
Transparency Coalition.ai
Opposition
None on file.
Analysis Prepared by: Slater Sharp / P. & C.P. / (916) 319-2200
Brian Contreras, “Tougher AI Policies Could Protect Taylor Swift—And Everyone Else—From Deepfakes,” Scientific American (Feb. 8. 2024) accessed at www.scientificamerican.com/article/tougher-ai-policies-could-protect-taylor-swift-and-everyone-else-from-deepfakes/. ↑
Natasha Singer, “Teen Girls Confront an Epidemic of Deepfake Nudes in Schools”, The New York Times (Apr. 8, 2024), https://www.nytimes.com/2024/04/08/technology/deepfake-ai-nudes-westfield-high-school.html. ↑
Mackenzie Tatananni, “ ‘Inappropriate images’ circulate at yet another California high school, as officials grapple with how to protect teens from AI porn created by classmates,” Daily Mail (Apr. 11, 2024) accessed at https://www.dailymail.co.uk/news/article-13295475/Inappropriate-images-California-Fairfax-High-School-AI-deepfake.html. ↑
Tim McNicholas, “New Jersey high school students accused of making AI-generated pornographic images of classmates,” CBS News (Nov. 2, 2023), https://www.cbsnews.com/newyork/news/westfield-high-school-ai-pornographic-images-students/; Lauraine Langreo, “Students Are Sharing Sexually Explicit ‘Deepfakes.’ Are Schools Prepared?” Ed Week (Sept. 26, 2024), https://www.edweek.org/leadership/studentsare-sharing-sexually-explicit-deepfakes-are-schools-prepared/2024/09; Gabrielle Hunt and Daryl Higgens “AI nudes of Victorian students were allegedly shared online. How can schools and parents respond to deepfake porn?,” The Guardian (June, 12, 2024),
https://www.theguardian.com/australia-news/article/2024/jun/12/ai-nudes-of-victorian-students-were-allegedly-shared-online-how-canschools-and-parents-respond-to-deepfake-porn. ↑
People of the State of California v. Sol Ecom, Inc, et al. (2024) Case No. CGC-24-617237, p. 2, https://www.sfcityattorney.org/wp-content/uploads/2024/08/2024-08-16-First-Amended-Complaint_Redacted.pdf ↑
Elizabeth Laird, Maddy Dwyer and Kristin Woelfel, “In Deep Trouble: Surfacing Tech-Powered Sexual Harassment in K-12 Schools,” Center for Democracy & Technology (Sept. 26, 2024), https://cdt.org/wp-content/uploads/2024/09/FINAL-UPDATED-CDT-2024-NCII-Polling-Slide-Deck. Pdf. ↑
Emily Flitter and Stacy Cowley, "Voice Deepfakes Are Coming for Your Bank Balance", New York Times (Aug. 30, 2023), www.nytimes.com/2023/08/30/business/voice-deepfakes-bank-scams.html. ↑
Harvey Kong, "‘Everyone looked real’: multinational firm’s Hong Kong office loses HK$200 million after scammers stage deepfake video meeting," South China Morning Post (Feb. 4, 2024), www.scmp.com/news/hong-kong/law-and-crime/article/3250851/everyone-looked-real-multinational-firms-hong-kong-office-loses-hk200-million-after-scammers-stage. ↑
Department of Justice Federal Bureau of Investigations, “Criminals Use Generative Artificial Intelligence to Facilitate Financial Fraud” (Dec. 3, 2024) , https://www.ic3.gov/PSA/2024/PSA241203. ↑
Curt Devine, Donie O'Sullivan, Sean Lyngass, "A fake recording of a candidate saying he’d rigged the election went viral. Experts say it’s only the beginning," CNN (Feb. 1, 2024), www.cnn.com/2024/02/01/politics/election-deepfake-threats-invs/index.html. ↑
Ken Bensinger, “Elon Musk Shares Manipulated Harris Video, in Seeming Violation of X’s Policies”, The New York Times (July 27, 2024), https://www.nytimes.com/2024/07/27/us/politics/elon-musk-kamala-harris-deepfake.html ↑
SENATE RULES COMMITTEE Office of Senate Floor Analyses (916) 651-4171 | SB 1000 |
THIRD READING
Bill No: SB 1000
Author: Becker (D)
Amended: 3/26/26
Vote: 27 - Urgency
SENATE PRIV., DIGITAL TECH. & CONS. PROT. COMMITTEE: 8-1, 4/13/26
AYES: Cabaldon, Gonzalez, McNerney, Ochoa Bogh, Padilla, Reyes, Umberg, Wiener
NOES: Jones
SENATE APPROPRIATIONS COMMITTEE: 5-0, 5/14/26
AYES: Cervantes, Cabaldon, Grayson, Richardson, Wahab
NO VOTE RECORDED: Seyarto, Dahle
SUBJECT: California AI Transparency Act
SOURCE: Author
DIGEST:
This bill makes a number of changes to the California AI Transparency (CAIT) Act, including eliminating the requirement to offer manifest disclosures, removing the user threshold on the definition of “covered provider,” and amending other definitions and exemptions.
ANALYSIS:
Existing law:
- Establishes the CAIT Act, which becomes operative, in part, on August 2, 2026, and requires certain “covered providers” to make an artificial intelligence (AI) detection tool available at no cost by which a person can assess whether content was created or altered by the provider’s GenAI system. (Business (Bus.) & Professions (Prof.) Code § 22757 et seq.)
- Requires a “covered provider,” a person that creates, codes, or otherwise produces a GenAI system that has over 1 million monthly visitors or users and is publicly accessible within the geographic boundaries of the state, to offer users the option to include in AI-generated image, video, or audio content created by its own generative AI system a manifest disclosure that meets specified criteria, including that it identifies the content as AI-generated content. (Bus. & Prof. Code § 22757.3(a).)
- Requires a covered provider to include in AI-generated image, audio, and video content created by its generative AI system a latent disclosure that is detectable by the tool specified above and is, to the extent technically feasible, permanent or extraordinarily difficult to remove. (Bus. & Prof. Code § 22757.3(b).)
- Prohibits a covered provider from doing any of the following in carrying out the duties above:
- Collect or retain personal information when a person utilizes the covered provider’s AI detection tool, except that it may collect and retain the contact information of a person who submitted feedback.
- Retain any content submitted to the AI detection tool for longer than is necessary to comply with this law. (Bus. & Prof. Code § 22757.2(c).)
- Requires a large online platform, starting January 1, 2027, to do one of the following:
- Detect whether any provenance data that is compliant with widely adopted specifications adopted by an established standards-setting body is embedded into or attached to content distributed on the large online platform.
- Provide a user interface to disclose the availability of system provenance data that reliably indicates that the content was generated or substantially altered by a GenAI system or captured by a capture device. The user interface shall make clearly and conspicuously available to users information sufficient to identify the content’s authenticity, origin, or history of modification, including specified information such as whether provenance data is available.
- Allow a user to inspect all available system provenance data that is compliant with widely adopted specifications adopted by an established standards-setting body in an easily accessible manner by any of several specified means. (Bus. & Prof. Code § 22757.3.1.)
- Provides that violators of the above provisions are liable for a civil penalty in the amount of $5,000 per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel. Each day in violation is deemed a discrete violation. (Bus. & Prof. Code § 22757.4.)
- Provides that the CAIT Act does not apply to any product, service, website, or application that provides exclusively non-user-generated video game, television, streaming, movie, or interactive experiences.
This bill:
- Removes the user threshold from the definition of “covered provider.”
- Removes the requirement for covered providers to offer users the option to include a manifest disclosure in content. Requires the latent disclosure to be compliant or interoperable with widely accepted industry standards.
- Redefines “provenance data” to mean information about the origin of a piece of content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
- Replaces references to “AI detection tool” with “disclosure verification tool.”
- Provides that the tool need not detect altered content if only a “minor modification,” which includes any of the following alterations:
- A change to brightness, contrast, or color.
- Sharpening.
- Saturating.
- Resizing.
- Scaling.
- Cropping.
- Format conversions.
- Resampling.
- Denoising and removal of background noise in audio.
- Permits the tool to output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content pursuant to a notice that does both of the following:
- Informs the user of the personal information that may be output by the tool.
- Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the file’s permanent digital footprint and cannot be retracted from copies already in circulation.
- Authorizes covered providers to impose reasonable limitations on access to the tool to prevent misuse of the tool for malicious purposes.
- Reworks the provisions regarding a users’ personal information to provide that a covered provider shall not collect, use, or retain personal information from a user of the covered provider’s AI disclosure verification tool or any content submitted to the disclosure verification tool beyond what is reasonably necessary for user authentication. A covered provider shall not make access to the covered provider’s GenAI system or disclosure verification tool contingent upon providing personal information beyond what is strictly necessary for the specified purposes.
- Reworks the third-party obligations to provide that a covered provider must require by contract that the third-party licensee ensures that the system includes provenance data that meets the specified criteria laid out in the CAIT Act to the extent it is technically feasible. It reduces the time within which a covered provider must revoke a license from 96 hours to 48 hours.
- Amends the exemption to provide that the Act does not apply to a product, service, website, or application that provides exclusively nonuser-generated videogame systems incapable of producing highly realistic videos or images that a reasonable person could confuse with reality.
- Makes other technical and clarifying changes to the CAIT Act.
- Includes an urgency clause.
Background
Certain forms of media – audio recordings, video recordings, and still images – can be powerful evidence of the truth. While such media have always been susceptible to some degree of manipulation, fakes were relatively easy to detect. The rapid advancement of AI technology, specifically the wide-scale introduction of GenAI models, has made it drastically cheaper and easier to produce synthetic content created, audio, images, text, and video recordings that are not real, but that are so realistic that they are virtually impossible to distinguish from authentic content.
Last session, the Legislature passed the California AI Transparency Act (CAIT), which, beginning August 2, 2026, regulates provenance data disclosure in AI-generated content. “Covered providers” are required to embed latent provenance disclosures in content generated by their GenAI systems and to offer manifest disclosures. Providers are required to make available an AI detection tool at no cost to the user that meets certain criteria, including that it allows a user to assess whether certain content was created or altered by the covered provider’s system.
This bill is a product of many stakeholder conversations that seeks to clean up the CAIT Act and ensure it accomplishes its intended goal of providing useful transparency to Californians. The bill is author-sponsored. It is supported by Transparency Coalition.AI and ForensicVB LLC. No timely opposition was received.
Comment
Last session, the Legislature responded to some of these issues by passing the CAIT Act, SB 942 (Becker, Ch. 291, Stats. 2024), which is set to become operative, in part, later this year. The CAIT Act imposes obligations on “covered providers,” persons that create, code, or otherwise produce a GenAI system that has over 1 million monthly visitors or users and is publicly accessible within the geographic boundaries of the state. It requires such providers to make an AI detection tool available at no cost by which a person can assess whether content was created or altered by the provider’s GenAI system.
The CAIT Act also regulates. Covered providers are required to include a latent disclosure in AI-generated image, video, or audio content that is created by their GenAI system that is detectable using the above tool, and that is, to the extent technically feasible, permanent or extraordinarily difficult to remove. This latent disclosure must identify the provider, the tool, and the time and date of the content’s creation or alteration. Covered providers are also required to provide users making such content with their system with the option to include a manifest disclosure that identifies it as AI-generated content.
Last year, AB 853 (Wicks, Chapter 674, Statutes of 2025) bolstered the CAIT Act by establishing similar transparency requirements on large online platforms, capture device manufacturers, and GenAI system hosting platforms.
This bill seeks to clean up the CAIT act and respond to some concerns raised by stakeholders. It includes a number of technical and clarifying amendments. Specifically, the bill removes the requirement for covered providers to provide the option of including a manifest disclosure in the relevant content. There were concerns that such perceptible labels would be easy to spoof and could lead to confusion for users.
This bill ensures that more minor AI alterations do not trigger the requirements of this bill. It also allows users to see personal information in provenance data if the user to whom the information pertains expressly consents to such information being included in the provenance, and is informed that such disclosures are irreversible once content is disseminated outside their control. This provides the option for users to include such information if, for instance, they are looking for attribution. This bill removes the user threshold for covered providers to ensure obligations for including disclosures also fall on those who download and modify open-source systems.
This bill also redefines “provenance data” to mean information about the origin of a piece of content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body. The goal is to ensure the whole Act, as amended by AB 853, is aligned, and that large online platforms are deciphering such information.
According to the author:
As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren’t yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g., file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established by SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren’t interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. In light of the low implementation cost of current provenance standards relative to costs of model development, this bill removes the user threshold from the definition of "covered provider” to expand obligations to any AI system which is publicly available within the state, and aligns the act's requirements to reflect the state of content provenance standards and international provenance disclosure regimes. This bill is critical to ensuring California’s content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from.
FISCAL EFFECT: Appropriation: No Fiscal Com.: Yes Local: No
According to the Senate Appropriations Committee:
- Unknown costs to the state funded trial court system (Trial Court Trust Fund, General Fund) to adjudicate civil actions. By expanding who may be considered a covered provider and, consequently, subject to civil penalties with statutory damages, this bill may lead to additional case filings. The fiscal impact of this bill to the courts will depend on many unknowns, including the number of cases filed and the factors unique to each case. An eight-hour court day costs approximately $10,500 in staff workload. While the courts are not funded on a workload basis, an increase in workload could result in delayed court services and would put pressure on the General Fund to fund additional staff and resources and to increase the amount appropriated to backfill for trial court operations.
- The Department of Justice (DOJ) does not anticipate a significant fiscal impact.
- The California Department of Technology (CDT) notes that it would not fall within the definition of covered provider under this bill; therefore, it anticipates only minor and absorbable costs to continue tracking developments in the AI space.
SUPPORT: (Verified 5/14/26)
ForensicVB LLC
Transparency Coalition.AI
OPPOSITION: (Verified 5/14/26)
None received
ARGUMENTS IN SUPPORT: ForensicVB LLC writes:
From an operational standpoint, manifest disclosure or visible labels introduce a significant risk of investigative de-prioritization. When content is labeled—accurately or not—as AI-generated, it can create uncertainty at the front end of an investigation, delay triage decisions, and divert limited law enforcement resources away from time-sensitive CSAM cases. Put more simply, it has been reported that pedophiles operating on the Dark Web are labeling real CSAM content as “AI-generated” in the expectation that investigators will see the label and focus their attention on other non-labeled content.
For these reasons, I believe SB 1000’s current reliance on latent, provenance-based disclosure frameworks reflects a more responsible and technically sound path forward, and is consistent with existing content-provenance approaches already in use, such as the Coalition for Content Provenance and Authenticity (C2PA) specifications. The same professional experience that has led me to oppose manifest disclosure requirements has also led me to support latent disclosure approaches as a more reliable, durable, and effective trust signal for consumers of digital content. When implemented correctly, latent disclosure mechanisms can support authenticity, attribution, and accountability without creating avoidable investigative friction or opportunities for misuse.
Prepared by: Christian Kurpiewski / P., D.T., & C.P. / (916) 651-1548
5/14/26 16:36:48
**** END ****
SENATE COMMITTEE ON
Senator Sabrina Cervantes, Chair
2025 - 2026 Regular Session
SB 1000 (Becker) - California AI Transparency Act
Version: | Policy Vote: |
Urgency: | Mandate: |
Hearing Date: | Consultant: |
Bill Summary:
SB 1000 makes a number of changes to the California AI Transparency (CAIT) Act.
Fiscal Impact:
- Unknown costs to the state funded trial court system (Trial Court Trust Fund, General Fund) to adjudicate civil actions. By expanding who may be considered a covered provider and, consequently, subject to civil penalties with statutory damages, this bill may lead to additional case filings. The fiscal impact of this bill to the courts will depend on many unknowns, including the number of cases filed and the factors unique to each case. An eight-hour court day costs approximately $10,500 in staff workload. While the courts are not funded on a workload basis, an increase in workload could result in delayed court services and would put pressure on the General Fund to fund additional staff and resources and to increase the amount appropriated to backfill for trial court operations.
- The Department of Justice (DOJ) does not anticipate a significant fiscal impact.
- The California Department of Technology (CDT) notes that it would not fall within the definition of covered provider under this bill; therefore, it anticipates only minor and absorbable costs to continue tracking developments in the AI space.
Background:
The CAIT Act, beginning August 2, 2026, regulates provenance data disclosure in AI-generated content. “Covered providers” are required to embed latent provenance disclosures in content generated by their generative AI (GenAI) systems and to offer manifest disclosures. Providers are required to make available an AI detection tool at no cost to the user that meets certain criteria, including that it allows a user to assess whether certain content was created or altered by the covered provider’s system.
This bill seeks to clean up the CAIT Act and ensure it accomplishes its intended goal of providing useful transparency to Californians.
Proposed Law:
- Removes the user threshold from the definition of “covered provider.”
- Removes the requirement for covered providers to offer users the option to include a manifest disclosure in content. Requires the latent disclosure to be compliant or interoperable with widely accepted industry standards.
- Redefines “provenance data” to mean information about the origin of a piece of content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
- Replaces references to “AI detection tool” with “disclosure verification tool.”
- Provides that the tool need not detect altered content if only a “minor modification,” which includes any of the following alterations:
- A change to brightness, contrast, or color.
- Sharpening.
- Saturating.
- Resizing.
- Scaling.
- Cropping.
- Format conversions.
- Resampling.
- Denoising and removal of background noise in audio.
- Permits the tool to output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content pursuant to a notice that does both of the following:
- Informs the user of the personal information that may be output by the tool.
- Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the file’s permanent digital footprint and cannot be retracted from copies already in circulation.
- Authorizes covered providers to impose reasonable limitations on access to the tool to prevent misuse of the tool for malicious purposes.
- Reworks the provisions regarding a users’ personal information to provide that a covered provider shall not collect, use, or retain personal information from a user of the covered provider’s AI disclosure verification tool or any content submitted to the disclosure verification tool beyond what is reasonably necessary for user authentication. A covered provider shall not make access to the covered provider’s GenAI system or disclosure verification tool contingent upon providing personal information beyond what is strictly necessary for the specified purposes.
- Reworks the third-party obligations to provide that a covered provider must require by contract that the third-party licensee ensures that the system includes provenance data that meets the specified criteria laid out in the CAIT Act to the extent it is technically feasible. It reduces the time within which a covered provider must revoke a license from 96 hours to 48 hours.
- Amends the exemption to provide that the Act does not apply to a product, service, website, or application that provides exclusively nonuser-generated videogame systems incapable of producing highly realistic videos or images that a reasonable person could confuse with reality.
- Makes other technical and clarifying changes to the CAIT Act.
- Includes an urgency clause and states the measure is necessary for establishing an ecosystem of information online that allows consumers to verify the provenance of content, including, but not limited to, whether or not it was generated or modified by a GenAI system is paramount to combating misinformation and disinformation as it relates to public health, election integrity, preventing fraud, extortion, or using generative artificial intelligence to create content for other malicious purposes.
- States legislative findings and declarations.
-- END --
SENATE PRIVACY, DIGITAL TECHNOLOGIES, AND CONSUMER PROTECTION COMMITTEE
Senator Christopher Cabaldon, Chair
2025-2026 Regular Session
Bill No: SB 1000 (Author:Becker)
Version: March 26, 2026
Hearing Date: April 13, 2026
Fiscal: Yes
Urgency: Yes
Consultant: CK
SUBJECT
California AI Transparency Act
DIGEST
This bill makes a number of changes to the California AI Transparency Act, including eliminating the requirement to offer manifest disclosures, removing the user threshold on the definition of “covered provider,” and amending other definitions and exemptions.
EXECUTIVE SUMMARY
Certain forms of media – audio recordings, video recordings, and still images – can be powerful evidence of the truth. While such media have always been susceptible to some degree of manipulation, fakes were relatively easy to detect. The rapid advancement of AI technology, specifically the wide-scale introduction of GenAI models, has made it drastically cheaper and easier to produce synthetic content created, audio, images, text, and video recordings that are not real, but that are so realistic that they are virtually impossible to distinguish from authentic content.
Last session, the Legislature passed the California AI Transparency Act (CAIT), which, beginning August 2, 2026, regulates provenance data disclosure in AI-generated content. “Covered providers” are required to embed latent provenance disclosures in content generated by their GenAI systems and to offer manifest disclosures. Providers are required to make available an AI detection tool at no cost to the user that meets certain criteria, including that it allows a user to assess whether certain content was created or altered by the covered provider’s system.
This bill is a product of many stakeholder conversations that seeks to clean up the CAIT Act and ensure it accomplishes its intended goal of providing useful transparency to Californians. The bill is author-sponsored. It is supported by Transparency Coalition.AI and ForensicVB LLC. No timely opposition was received by the Committee.
PROPOSED CHANGES TO THE LAW
Existing law:
- Establishes the California AI Transparency Act, which becomes operative, in part, on August 2, 2026, and requires certain “covered providers” to make an AI detection tool available at no cost by which a person can assess whether content was created or altered by the provider’s GenAI system. (Bus. & Prof. Code § 22757 et seq.)
- Requires a “covered provider,” a person that creates, codes, or otherwise produces a GenAI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state, to offer users the option to include in AI-generated image, video, or audio content created by its own generative AI system a manifest disclosure that meets specified criteria, including that it identifies the content as AI-generated content. (Bus. & Prof. Code § 22757.3(a).)
- Requires a covered provider to include in AI-generated image, audio, and video content created by its generative AI system a latent disclosure that is detectable by the tool specified above and is, to the extent technically feasible, permanent or extraordinarily difficult to remove. (Bus. & Prof. Code § 22757.3(b).)
- Prohibits a covered provider from doing any of the following in carrying out the duties above:
- Collect or retain personal information when a person utilizes the covered provider’s AI detection tool, except that it may collect and retain the contact information of a person who submitted feedback.
- Retain any content submitted to the AI detection tool for longer than is necessary to comply with this law. (Bus. & Prof. Code § 22757.2(c).)
- Requires a large online platform, starting January 1, 2027, to do one of the following:
- Detect whether any provenance data that is compliant with widely adopted specifications adopted by an established standards-setting body is embedded into or attached to content distributed on the large online platform.
- Provide a user interface to disclose the availability of system provenance data that reliably indicates that the content was generated or substantially altered by a GenAI system or captured by a capture device. The user interface shall make clearly and conspicuously available to users information sufficient to identify the content’s authenticity, origin, or history of modification, including specified information such as whether provenance data is available.
- Allow a user to inspect all available system provenance data that is compliant with widely adopted specifications adopted by an established standards-setting body in an easily accessible manner by any of several specified means. (Bus. & Prof. Code § 22757.3.1.)
- Provides that violators of the above provisions are liable for a civil penalty in the amount of $5,000 per violation to be collected in a civil action filed by the Attorney General, a city attorney, or a county counsel. Each day in violation is deemed a discrete violation. (Bus. & Prof. Code § 22757.4.)
- Provides that the CAIT Act does not apply to any product, service, website, or application that provides exclusively non-user-generated video game, television, streaming, movie, or interactive experiences.
This bill:
- Removes the user threshold from the definition of “covered provider.”
- Removes the requirement for covered providers to offer users the option to include a manifest disclosure in content. Requires the latent disclosure to be compliant or interoperable with widely accepted industry standards.
- Redefines “provenance data” to mean information about the origin of a piece of content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body.
- Replaces references to “AI detection tool” with “disclosure verification tool.”
- Provides that the tool need not detect altered content if only a “minor modification,” which includes any of the following alterations:
- A change to brightness, contrast, or color.
- Sharpening.
- Saturating.
- Resizing.
- Scaling.
- Cropping.
- Format conversions.
- Resampling.
- Denoising and removal of background noise in audio.
- Permits the tool to output personal information that is detected in the content if the user to whom the personal information pertains expressly consents, clearly and conspicuously in plain language, to including personal information specified by the user in the content pursuant to a notice that does both of the following:
- Informs the user of the personal information that may be output by the tool.
- Informs the user that once personal information is embedded into provenance data and exported, the information becomes part of the file’s permanent digital footprint and cannot be retracted from copies already in circulation.
- Authorizes covered providers to impose reasonable limitations on access to the tool to prevent misuse of the tool for malicious purposes.
- Reworks the provisions regarding a users’ personal information to provide that a covered provider shall not collect, use, or retain personal information from a user of the covered provider’s AI disclosure verification tool or any content submitted to the disclosure verification tool beyond what is reasonably necessary for user authentication. A covered provider shall not make access to the covered provider’s GenAI system or disclosure verification tool contingent upon providing personal information beyond what is strictly necessary for the specified purposes.
- Reworks the third-party obligations to provide that a covered provider must require by contract that the third-party licensee ensures that the system includes provenance data that meets the specified criteria laid out in the CAIT Act to the extent it is technically feasible. It reduces the time within which a covered provider must revoke a license from 96 hours to 48 hours.
- Amends the exemption to provide that the Act does not apply to a product, service, website, or application that provides exclusively nonuser-generated videogame systems incapable of producing highly realistic videos or images that a reasonable person could confuse with reality.
- Makes other technical and clarifying changes to the CAIT Act.
- Includes an urgency clause.
COMMENTS
- Blurring reality: AI-generated content
GenAI can create new content, including text, images, code, or music, by learning from existing data. GenAI models can produce realistic and novel artifacts that resemble the data they were trained on, but do not copy it. The world has been in awe of the powers of this new technology, but the capabilities of these advanced systems lead to a blurring between reality and fiction. The National Institute for Standards and Technology frames the issue presented:
In recent years, the quality of digital content generated by artificial intelligence (AI) has advanced considerably across various modalities, including image, video, audio, and text. This surge in generative AI capability presents both opportunities and challenges—generative AI has facilitated creative expression and production, enabling artists, designers, and writers to create digital content at a much faster pace, but has also raised concerns regarding the authenticity and integrity of digital media, especially as it has become increasingly difficult to distinguish AI-generated content from human-generated content.[1]
The relevant issue here is not about the growing incidence of synthetic content alone, but about the quickly blurring line between what is real and what is not. As one recent article phrased it, there is a “New Authenticity Crisis”:
OpenAI launched Sora 2 recently, an app that generates photorealistic video from text prompts. Within a week, Hollywood talent agencies called it exploitation. The Motion Picture Association objected to its copyright policies. Then came the flood of videos featuring copyrighted characters in unauthorized scenes. Despite being invite-only, the app hit one million downloads in five days. Faster than ChatGPT’s launch. Social media is now overwhelmed with synthetic clips so convincing that Zelda Williams, daughter of Robin Williams, publicly asked strangers to stop sending her AI-generated videos of her deceased father.
This is not an isolated technology story. It represents the acceleration of something larger happening across business and wider society.[2]
- Cleaning up CAIT
Last session, the Legislature responded to some of these issues by passing the California AI Transparency Act (CAIT Act), SB 942 (Becker, Ch. 291, Stats. 2024), which is set to become operative, in part, later this year. The CAIT Act imposes obligations on “covered providers,” persons that create, code, or otherwise produce a GenAI system that has over 1,000,000 monthly visitors or users and is publicly accessible within the geographic boundaries of the state. It requires such providers to make an AI detection tool available at no cost by which a person can assess whether content was created or altered by the provider’s GenAI system.
The CAIT Act also regulates. Covered providers are required to include a latent disclosure in AI-generated image, video, or audio content that is created by their GenAI system that is detectable using the above tool, and that is, to the extent technically feasible, permanent or extraordinarily difficult to remove. This latent disclosure must identify the provider, the tool, and the time and date of the content’s creation or alteration. Covered providers are also required to provide users making such content with their system with the option to include a manifest disclosure that identifies it as AI-generated content.
Last year, AB 853 (Wicks, Ch. 674, Stats. 2025) bolstered the CAIT Act by establishing similar transparency requirements on large online platforms, capture device manufacturers, and GenAI system hosting platforms.
This bill seeks to clean up the Act and respond to some concerns raised by stakeholders. It includes a number of technical and clarifying amendments. Specifically, the bill removes the requirement for covered providers to provide the option of including a manifest disclosure in the relevant content. There were concerns that such perceptible labels would be easy to spoof and could lead to confusion for users.
The bill ensures that more minor AI alterations do not trigger the requirements of the bill. It also allows users to see personal information in provenance data if the user to whom the information pertains expressly consents to such information being included in the provenance, and is informed that such disclosures are irreversible once content is disseminated outside their control. This provides the option for users to include such information if, for instance, they are looking for attribution. The bill removes the user threshold for covered providers to ensure obligations for including disclosures also fall on those who download and modify open-source systems.
The bill also redefines “provenance data” to mean information about the origin of a piece of content and the history of modifications to the content that is in a format that is compliant, or interoperable with, widely adopted specifications adopted by an established standards-setting body. The goal is to ensure the whole Act, as amended by AB 853, is aligned, and that large online platforms are deciphering such information.
According to the author:
As AI technology advances, distinguishing between human and machine-generated content becomes increasingly challenging. This ambiguity poses significant risks to Californians, exacerbating problems of disinformation, harassment, and fraud while threatening the integrity of the information environment our democracy and economy depend on. In 2024, the legislature passed SB 942, the first bill in the nation to establish disclosure requirements for synthetic content. Since then, content provenance technology has developed in such a way that the leading technologies for embedding provenance are not accurately described by the law. Additionally, some methods of marking content are still in their nascency, and aren’t yet robust enough to withstand adversarial exploitation or certain kinds of transformations (e.g., file compression, spoofing attacks, added noise to an image, or file type conversions). The original act also did not address the risk that visible labels create a binary signal implying that unlabeled content is authentic and labeled content is suspect - regardless of whether either inference is warranted. Furthermore, obligations for developers established by SB 942 do not require that the information embedded in AI generated or modified content be consistent - or interoperable with - widely adopted standards. Last year, AB 853 (Wicks) established obligations for large online platforms to read provenance data, but only if it was compliant with such standards. This gap risks establishing a fragmented ecosystem where new methods of provenance disclosure that aren’t interoperable with current standards are not read by large online platforms, and never subsequently relayed to users. In light of the low implementation cost of current provenance standards relative to costs of model development, this bill removes the user threshold from the definition of "covered provider” to expand obligations to any AI system which is publicly available within the state, and aligns the act's requirements to reflect the state of content provenance standards and international provenance disclosure regimes. This bill is critical to ensuring California’s content labeling laws are effective at providing consumers with consistent information about where the content they see online comes from.
Writing in support, ForensicVB LLC makes the case for relying solely on latent disclosures:
From an operational standpoint, manifest disclosure or visible labels introduce a significant risk of investigative de-prioritization. When content is labeled—accurately or not—as AI-generated, it can create uncertainty at the front end of an investigation, delay triage decisions, and divert limited law enforcement resources away from time-sensitive CSAM cases. Put more simply, it has been reported that pedophiles operating on the Dark Web are labeling real CSAM content as “AI-generated” in the expectation that investigators will see the label and focus their attention on other non-labeled content.
For these reasons, I believe SB 1000’s current reliance on latent, provenance-based disclosure frameworks reflects a more responsible and technically sound path forward, and is consistent with existing content-provenance approaches already in use, such as the Coalition for Content Provenance and Authenticity (C2PA) specifications. The same professional experience that has led me to oppose manifest disclosure requirements has also led me to support latent disclosure approaches as a more reliable, durable, and effective trust signal for consumers of digital content. When implemented correctly, latent disclosure mechanisms can support authenticity, attribution, and accountability without creating avoidable investigative friction or opportunities for misuse.
The California Initiative for Technology and Democracy (CITED) writes in a support if amended position:
While CITED applauds the improvements outlined above, we believe there are still continued refinements that deserve attention. For instance, the bill excludes the listing of “minor modifications” in provenance information, a term defined by a listing of editing techniques. While this change generally aligns with CITED’s goals of making provenance information useful and accessible to users, we hope additional specificity is added to the description of the editing techniques to avoid ambiguity.
Additionally, CITED objects to the elimination of the requirement for Gen AI Providers to provide users with the option to place manifest disclosures on the bill. Manifest disclosures remain the most accessible way a consumer can quickly discern the origins of digital content. We are encouraged by the Washington Legislature’s adoption of HB 1170, which mirrors the requirements of the California AI Transparency Act, including the aforementioned manifest disclosures. We maintain that the utility of a visible disclosure outweighs the potential risk of misleading the public in the vast majority of use cases.
SUPPORT
ForensicVB LLC
Transparency Coalition.AI
OPPOSITION
None received
RELATED LEGISLATION
SB 1050 (Ashby, 2026) makes it an unlawful practice for a person, in connection with the creation or dissemination of an advertisement in this state, to use or cause to be used a synthetic performer, as defined, without a clear and conspicuous disclosure that the performer is synthetic. SB 1050 is currently in the Senate Judiciary Committee.
SB 1146 (Gonzalez, 2026) requires an advertisement to include a disclosure, as applicable, when it depicts a digitally altered or generated person representing themselves to be, or “identifiably depicting” a person as, a health care provider to promote the sale of a health-related consumer product or service, except as provided. SB 1146 is currently in the Senate Judiciary Committee.
AB 853 (Wicks, Ch. 674, Stats. 2025) See Comment 2.
SB 942 (Becker, Ch. 291, Stats. 2024) See Comment 2.
**************
NIST, GenAI: Text 2026, https://ai-challenges.nist.gov/text-2026. All internet citations are current as of March 23, 2026. ↑
Prashant Saxena, Authenticity in the Age of AI (December 22, 2025) California Management Review, https://cmr.berkeley.edu/assets/documents/pdf/2025-12-authenticity-in-the-age-of-ai.pdf. ↑
Discussed in Hearing
Senate Floor
Assembly Floor
Assembly Standing Committee on Appropriations
Senate Floor
Senate Standing Committee on Privacy, Digital Technologies, and Consumer Protection
Bill Author
Bill Co-Author(s):