The rapid adoption of artificial intelligence (“AI”), including generative AI, has prompted lawmakers to grapple with how existing legal frameworks apply to the technology and where new rules may be needed. In the U.S., federal and state lawmakers are targeting issues that range from the use of copyrighted works to train AI models and the creation of digital replicas to AI-generated content, algorithmic decision-making and pricing, and the safety and governance of AI systems.

Against this evolving backdrop, TFL is tracking key domestic AI legislation across AI training, copyright & intellectual property; digital replicas, likeness & synthetic performers; AI-generated content, disclosure & provenance; algorithmic decision-making, pricing & consumer protection; and AI systems, safety & governance. The tracker is not exhaustive and focuses on legislation with particular implications for brands, retailers, online platforms, technology companies, and intellectual property owners.

LAST UPDATED SEPT. 30, 2026 · UPDATED REGULARLY · U.S. AI LEGISLATION

AI Training & Intellectual Property

June 2, 2026: Creative Rights Ensuring Artists’ Technique and Originality are Reserved Act (CREATOR Act) (H.R. 9112)

Introduced: June 2, 2026 by Rep. Beth Van Duyne (R-TX), with Reps. Yvette Clarke (D-NY) and Valerie Foushee (D-NC).

Snapshot: The CREATOR Act would create a new federal intellectual property right allowing visual artists and other right holders to control the commercial exploitation or public distribution of AI-generated works that constitute a “stylistic impersonation” of a visual artist.

Key Provisions: The bill would establish a federal right, distinct from copyright and trademark, covering deliberate stylistic impersonation of a visual artist. Liability could arise from knowingly selling, licensing, publicly distributing, or otherwise commercially exploiting such works without authorization, as well as from developing and expressly marketing certain products or services intentionally configured to generate stylistic impersonations. The legislation includes protections for general-purpose AI systems, exclusions for commentary, criticism, scholarship, research, teaching, parody, satire, news reporting, and other specified uses, and a notice-and-removal safe harbor for online services. Right holders could seek injunctive relief, actual damages and profits, or statutory damages.

Potential Implications: The proposal would create a new form of protection for visual artists beyond existing copyright and trademark law. For designers, illustrators, photographers, graphic artists, and other visual creators, the legislation could provide a new basis for challenging AI-generated works that deliberately emulate distinctive visual characteristics associated with their work, while creating corresponding compliance considerations for AI developers and online platforms.

Status: Pending. The bill was referred to the House Committee on the Judiciary on June 2, 2026.

Jan. 22, 2026: Transparency and Responsibility for AI Networks Act (TRAIN Act) (H.R. 7209)

Introduced: Jan. 22, 2026 by Reps. Madeleine Dean (D-PA) and Nathaniel Moran (R-TX).

Snapshot: The Transparency and Responsibility for AI Networks (“TRAIN”) Act would establish a mechanism for copyright owners to determine whether their copyrighted works were used to train generative AI models by enabling them to obtain certain training records from AI developers.

Key Provisions: Copyright owners with a good-faith belief that their works were used to train a generative AI model could request an administrative subpoena requiring a developer to disclose specified training records. The subpoena would be limited to works owned by the requester, and recipients would be subject to confidentiality requirements. If a developer fails to comply with a valid subpoena, the legislation would create a rebuttable presumption that the copyrighted material was used in training. Courts could impose sanctions for bad-faith requests.

Potential Implications: The bill addresses one of the central challenges facing copyright owners seeking to enforce their rights against AI developers: determining whether their works are contained in datasets that are generally not publicly accessible. For brands, publishers, photographers, designers, and other rights holders, the proposed subpoena mechanism could provide a new route for obtaining information needed to assess potential copyright claims.

Status: Pending. The measure follows an earlier version of the TRAIN Act introduced by Sen. Peter Welch (D-VT) in November 2024, which expired at the end of the 118th Congress. 

Mar. 20, 2025: Arkansas – Ownership of Model Training and Content Generated by a Generative AI Tool (Act 927)

Introduced: Mar. 20, 2025 as H.B. 1876; signed into law May 2, 2025.

Snapshot: Arkansas Act 927 addresses ownership of certain inputs, outputs, and trained models associated with the use of generative AI tools, providing that a person who supplies data or other input generally owns the resulting content or trained model, subject to existing intellectual property rights.

Key Provisions: The law provides that an individual who provides input or data to a generative AI tool owns the resulting content or trained model, provided that doing so does not infringe existing intellectual property rights. Where the relevant activity occurs within the scope and direction of employment, the law provides for ownership by the employer.

Potential Implications: The law represents an unusual state-level attempt to address ownership questions surrounding AI-generated content and model training. Its provisions may be relevant to companies developing or commissioning AI-generated creative assets and to employment relationships in which workers use generative AI tools in the course of their work.

Status: Enacted May 2, 2025 as Act 927. 

March 6, 2025: New York – Artificial Intelligence Training Data Transparency Act (A.6578-B/S.6955-A)

Introduced: March 6, 2025 by Assemblymember Alex Bores (D-NY), with a Senate companion introduced by Sen. Andrew Gounardes (D-NY) on March 27, 2025.

Snapshot: The Artificial Intelligence Training Data Transparency Act would require developers of generative AI models and services made publicly available to New Yorkers to disclose information about the datasets used to train those systems.

Key Provisions: Developers would be required to post documentation on their websites providing a high-level summary of training datasets, including their sources or owners, the approximate number and types of data points, whether datasets contain data protected by copyright, trademark, or patent, whether datasets were purchased or licensed, whether they include personal or aggregate consumer information, how they were processed or modified, the periods during which data was collected and first used, and whether synthetic data was used. The requirement would apply to covered models or services, or substantial modifications to them, released on or after January 1, 2022 and made publicly available to New Yorkers. The bill includes exemptions for certain aviation and national security, military, or defense systems.

Potential Implications: The measure would give rights holders, consumers, and other stakeholders greater visibility into the materials used to develop generative AI systems. For brands, publishers, designers, photographers, and other intellectual property owners, the disclosures could provide information relevant to assessing whether protected content may have been incorporated into AI training datasets, although the legislation would not itself establish liability for the use of that material.

Status: Passed by the Legislature. The Assembly passed A.6578-B on May 5, 2026, and the Senate passed it on June 4, 2026. As of September 11, 2026, the measure has not been delivered to or signed by the governor.

Feb. 13, 2025: Pennsylvania – Clarifying AI & Copyright (H.R. 81)

Introduced: Feb. 13, 2025 by Rep. Kristine Howard.

Snapshot: H.R. 81 is a Pennsylvania House resolution calling on Congress to clarify the application of federal copyright law to AI-generated works and the use of copyrighted material to train AI systems.

Key Provisions: The resolution urges Congress to amend federal copyright law to clarify that works generated predominantly by AI are not eligible for copyright protection, that copyright protection is limited to human-authored works, and that scraping copyrighted works for AI model training does not constitute fair use.

Potential Implications: While the resolution would not itself alter copyright law, it reflects growing pressure on federal lawmakers to resolve questions surrounding human authorship, AI-generated works, and whether the unauthorized use of copyrighted works for AI training qualifies as fair use.

Status: Pending. 

Feb. 4, 2025: California – Generative AI: Training Data (AB 412)

Introduced: Feb. 4, 2025 by Assemblymember Rebecca Bauer-Kahan.

Snapshot: AB 412 would have required developers of generative AI models available in California to maintain records concerning copyrighted materials used in model training and provide information about those materials to copyright owners upon request.

Key Provisions: Developers would have been required to document copyrighted materials used to train their models and identify the copyright owners. Upon receiving a request from a rights holder, a developer would have had seven days to provide a list of relevant training materials or certify that the copyright owner’s works were not used. The proposal also contemplated civil actions for non-compliance.

Potential Implications: The bill sought to address the same underlying transparency problem as the federal TRAIN Act but would have imposed affirmative recordkeeping and disclosure obligations on developers rather than relying on a subpoena mechanism.

Status: Did not advance out of committee during the 2025 legislative year; not enacted. 

Jan. 31, 2024: California Generative Artificial Intelligence Training Data Transparency Act (AB 2013)

Introduced: Jan. 31, 2024 by Assemblymember Jacqui Irwin; signed into law Sept. 28, 2024.

Snapshot: The Generative Artificial Intelligence Training Data Transparency Act requires developers of generative AI systems or services made available to Californians to publicly disclose specified information about the data used to train those systems.

Key Provisions: Covered developers must disclose high-level information concerning their training datasets, including the sources or owners of the data; the types and volume of data used; how the data was collected and processed; whether copyrighted or licensed material is included; whether the datasets contain personal information; the period during which the data was collected; and whether synthetic data was used.

The law applies to generative AI systems or services released or substantially modified on or after Jan. 1, 2022, subject to specified exemptions. It does not require developers to disclose the training datasets themselves.

Potential Implications: The law introduces greater transparency into an area that has become central to copyright litigation and licensing disputes involving generative AI. For copyright owners, the disclosures may provide additional information about the types and sources of works used in model development, while AI developers face new obligations to document and publicly describe their training practices.

Status: Enacted; effective Jan. 1, 2026. 

Apr. 9, 2024: Generative AI Copyright Disclosure Act (H.R. 7913)

Introduced: Apr. 9, 2024 by Rep. Adam Schiff (D-CA).

Snapshot: The Generative AI Copyright Disclosure Act would require companies developing generative AI systems to notify the U.S. Copyright Office about copyrighted works used in the datasets employed to train those systems.

Key Provisions: Developers would be required to submit notice to the Register of Copyrights identifying copyrighted works used in building or altering a generative AI system’s training dataset before releasing the system to consumers. The disclosure obligation would also apply retroactively to previously released generative AI systems.

Potential Implications: The proposal would establish a federal disclosure regime around copyrighted training material, potentially giving rights holders greater visibility into whether their works were used in model development and creating new compliance obligations for AI developers.

Status: Introduced in the 118th Congress; did not become law before the end of the congressional session. 

Mar. 19, 2024: AI CONSENT Act (S.3975)

Introduced: Mar. 19, 2024 by Sen. Peter Welch (D-VT).

Snapshot: The AI CONSENT Act would require online platforms to obtain consumers’ express informed consent before using their personal data to train AI models.

Key Provisions: Platforms would be required to obtain affirmative consent before using covered personal information for AI training. Failure to obtain consent would constitute an unfair or deceptive act or practice subject to Federal Trade Commission enforcement. The legislation would also direct the FTC to study the effectiveness of data de-identification techniques in light of advances in AI.

Potential Implications: Unlike copyright-focused training-data proposals, the AI CONSENT Act approaches model training through privacy and consumer-protection law, potentially limiting companies’ ability to repurpose user data for AI development without explicit permission.

Status: Introduced in the 118th Congress; did not become law before the end of the congressional session. 

Dec. 22, 2023: AI Foundation Model Transparency Act of 2023 (H.R. 6881)

Introduced: Dec. 22, 2023 by Reps. Anna Eshoo (D-CA) and Don Beyer (D-VA).

Snapshot: The AI Foundation Model Transparency Act would have directed the Federal Trade Commission, in consultation with the National Institute of Standards and Technology and Office of Science and Technology Policy, to establish transparency standards for high-impact foundation models.

Key Provisions: The contemplated disclosures would include information concerning training data, how models were trained, and whether user data was collected during inference.

Potential Implications: The proposal reflected an early federal effort to establish standardized transparency requirements for foundation models, including visibility into training data and data-collection practices that have subsequently become central issues in AI regulation.

Status: Expired at the end of the 118th Congress.

Digital Replicas, Likeness & Synthetic Performers

June 9, 2025: New York – Synthetic Performers in Advertising (S.8420/A.8887)

Introduced: June 9, 2025 by Sen. Michael Gianaris; Assembly companion A.8887 introduced June 9, 2025.

Snapshot: The legislation requires advertisers to disclose the use of AI-generated “synthetic performers” in commercial advertisements, creating a new transparency requirement for brands and other businesses using virtual or AI-generated people in advertising.

Key Provisions: The law defines a “synthetic performer” as a digital asset created, reproduced, or modified using generative AI or another software algorithm that gives the impression of a human performance but is not recognizable as an identifiable natural performer. Advertisers that knowingly use a synthetic performer must conspicuously disclose that use. The law provides for civil penalties of $1,000 for a first violation and $5,000 for subsequent violations and includes exemptions for certain advertisements for expressive works, audio advertisements, and uses of AI solely for language translation. 

Potential Implications: The law is particularly relevant for fashion, beauty, and retail companies experimenting with AI-generated models and other synthetic talent in campaigns. It establishes a direct disclosure obligation around the use of wholly synthetic performers even where the AI-generated person does not replicate an identifiable individual.

Status: Signed into law Dec. 11, 2025 as Chapter 617 of the Laws of 2025. This entry consolidates the earlier proposed and enacted versions previously listed separately in the tracker.

Apr. 9, 2025: NO FAKES Act of 2025 (S.1367/H.R.2794)

Introduced: Apr. 9, 2025 in the Senate and House.

Snapshot: The Nurture Originals, Foster Art, and Keep Entertainment Safe (“NO FAKES”) Act would establish a federal right protecting individuals against unauthorized digital replicas of their voices and visual likenesses. The legislation is aimed at AI-generated and other highly realistic digital replicas that make an identifiable person appear to have said or done something they did not. 

Key Provisions: The bill would establish an individual right in voice and visual likeness and impose liability for producing unauthorized digital replicas and, subject to specified conditions, publishing, reproducing, displaying, distributing, transmitting, or otherwise making them available to the public. The legislation includes exceptions and limitations intended to protect certain news, documentary, commentary, parody, satire, and other expressive uses and establishes mechanisms governing online-service-provider liability.

Potential Implications: For fashion and retail companies, the legislation could affect the creation and use of AI-generated advertising, virtual models, digital campaigns, endorsements, and other content that replicates identifiable individuals. Brands, agencies, platforms, and technology providers would need to assess whether appropriate authorization exists before creating or commercially using digital replicas of models, celebrities, creators, and other individuals.

Status: Pending in the 119th Congress. The House version was referred to the House Judiciary Committee on Apr. 9, 2025. 

Dec. 13, 2024: New York Digital Replica Contracts Act (S.7676-B/A.8138-B)

Introduced: Oct. 3, 2023 by Sen. Jessica Ramos; signed into law Dec. 13, 2024.

Snapshot: New York’s Digital Replica Contracts Act establishes requirements governing contractual provisions that authorize the creation and use of digital replicas of performers.

Key Provisions: The law limits the enforceability of certain contractual provisions involving digital replicas where the provision does not include a reasonably specific description of the intended uses of the replica and the performer was not represented by legal counsel or a labor organization in negotiating the agreement.

Potential Implications: The law establishes guardrails around how companies obtain contractual rights to digitally replicate performers, making specificity and representation important considerations when negotiating agreements involving AI-generated versions of an individual’s voice or likeness.

Status: Enacted; signed into law Dec. 13, 2024.

Dec. 11, 2025: New York Posthumous Right of Publicity (S.8391/A.8882)

Introduced: 2025 by Sen. Michael Gianaris and Assemblymember Tony Simone.

Snapshot: The legislation expands New York’s posthumous right of publicity protections, including protections against unauthorized AI-generated digital replicas of deceased performers.

Key Provisions: The law amends New York’s definitions of “deceased performer,” “deceased personality,” and “digital replica” and restricts specified unauthorized uses of a deceased performer’s digital replica. It preserves exceptions for certain expressive uses, including parody, satire, criticism, and commentary. 

Potential Implications: The law expands the legal risks associated with using AI to recreate deceased performers and other personalities, with implications for entertainment companies, advertisers, brands, platforms, and others seeking to commercially deploy digital likenesses.

Status: Enacted; signed into law Dec. 11, 2025.

Jan. 10, 2024: No AI FRAUD Act (H.R. 6943)

Introduced: Jan. 10, 2024 by Rep. María Elvira Salazar (R-FL).

Snapshot: The No Artificial Intelligence Fake Replicas And Unauthorized Duplications (“No AI FRAUD”) Act proposed a federal property right in an individual’s likeness and voice in response to the increasing use of AI and deepfake technologies to create unauthorized replicas. 

Key Provisions: The bill would have established individual property rights in likeness and voice and provided a federal cause of action for certain unauthorized uses. The proposed rights would have been descendible and licensable, subject to limitations intended to account for First Amendment and other expressive interests.

Potential Implications: The proposal would have created a federal framework for rights that have traditionally been governed largely by state right-of-publicity law. For brands and advertisers, it could have increased the legal risk associated with creating or using AI-generated replicas of models, celebrities, influencers, and other identifiable individuals without authorization.

Status: Introduced in the 118th Congress and referred to the House Judiciary Committee; it did not become law before the end of the congressional session. 

Jan. 10, 2024: Tennessee – Ensuring Likeness Voice and Image Security (ELVIS) Act

Introduced: Jan. 10, 2024; signed into law March 21, 2024.

Snapshot: Tennessee’s ELVIS Act expands the state’s existing right-of-publicity protections to expressly protect an individual’s voice from unauthorized AI-generated and other uses.

Key Provisions: The law creates liability for certain unauthorized uses of an individual’s name, photograph, voice, or likeness and addresses the use of technology to produce or reproduce an individual’s voice or likeness without authorization. It also provides causes of action relating to the distribution or availability of technology whose primary purpose or function is producing unauthorized replicas, subject to statutory exceptions.

Potential Implications: Although closely associated with the music industry, the law has broader implications for advertising, endorsements, influencer marketing, and other commercial uses of identity. Fashion and retail companies using AI to replicate recognizable voices or likenesses may face liability even where the underlying content is entirely synthetic.

Status: Signed into law March 21, 2024; effective July 1, 2024.

Sept. 12, 2023: No Fakes Act of 2023

Introduced: Sept. 12, 2023 as a federal discussion draft.

Snapshot: The original NO FAKES proposal sought to create a federal right protecting individuals from unauthorized digital replicas of their voice or likeness.

Key Provisions: The proposal contemplated liability for producing or distributing unauthorized digital replicas while establishing exceptions for specified expressive and public-interest uses.

Potential Implications: The proposal marked an early effort to create a uniform federal digital-replica right in response to generative AI, addressing an area otherwise governed largely by a patchwork of state right-of-publicity laws.

Status: Superseded by the NO FAKES Act of 2025, which is tracked above.

AI-Generated Content, Disclosure & Provenance

July 2, 2026: Spot the Fakes Act (H.R. 9578)

Introduced: July 2, 2026 by Rep. Josh Gottheimer (D-NJ), with Reps. Tom Kean Jr. (R-NJ) and Sam Liccardo (D-CA).

Snapshot: The Spot the Fakes Act would require AI-generated content to be labeled as AI-generated through metadata or another technological means that provides indelible verification of provenance.

Key Provisions: A person that owns or controls AI-generated content would be required to label the content as AI-generated within its metadata or through another technological means determined appropriate by the National Institute of Standards and Technology. The requirement would apply to AI-generated video, images, audio, text, and other media. The Federal Trade Commission, in consultation with NIST, would establish the required method of disclosure, and violations would be treated as unfair or deceptive acts or practices under the FTC Act.

Potential Implications: The bill could create a federal provenance requirement extending across a broad range of AI-generated commercial and creative content. For brands, advertisers, retailers, platforms, and content creators, it could require systems for preserving and communicating AI-origin information as content moves through advertising, social media, e-commerce, and other digital channels.

Status: Pending. The bill was referred to the House Committee on Energy and Commerce on July 2, 2026.

May 19, 2026: Protecting Consumers from Deceptive AI Act (H.R. 8893)

Introduced: May 19, 2026 by Rep. Valerie Foushee (D-NC), with Reps. James Moylan (R-GU), Don Beyer (D-VA), and Brian Fitzpatrick (R-PA).

Snapshot: The Protecting Consumers from Deceptive AI Act would direct the National Institute of Standards and Technology to establish task forces to support the development of technical standards and guidelines for identifying and labeling content created or substantially modified by generative AI.

Key Provisions: The bill calls for NIST-led work on technical standards and guidelines for content provenance metadata, watermarking, digital fingerprinting, and other methods of identifying AI-generated or AI-modified content. It also contemplates interoperable approaches that could help online platforms identify, maintain, interpret, and display provenance information associated with audio, visual, and text-based content.

Potential Implications: Unlike legislation that would directly mandate AI-content labels, the bill focuses on the technical infrastructure needed to support identification, provenance, and labeling systems. The resulting standards could influence how brands, platforms, publishers, and other digital-content businesses authenticate and disclose AI-generated material across services.

Status: Pending. The House Committee on Science, Space, and Technology favorably reported the bill to the House on June 25, 2026 by a 35-0 vote, after adopting an amendment in the nature of a substitute and additional amendments.

Feb. 2, 2026: New York FAIR News Act (S.8451-B/A.8962-B)

Introduced: Feb. 2, 2026 by Sen. Patricia Fahy and Assemblymember Nily Rozic.

Snapshot: The New York Fundamental Artificial Intelligence Requirements in News (“FAIR News”) Act would establish disclosure, human-review, and workplace requirements governing the use of generative AI by news organizations operating in New York.

Key Provisions: News media employers would be required to disclose to workers when and how generative AI is used in connection with content creation. News content substantially created using generative AI would generally require a conspicuous disclosure to consumers, and AI-generated content would be subject to human review before publication. The legislation also includes protections concerning the use of journalists’ work product to train generative AI systems. 

Potential Implications: The proposal is particularly relevant to publishers and other media companies using AI in editorial workflows, but it also reflects a broader regulatory move toward requiring businesses to tell consumers when content has been generated by AI and to maintain human oversight of automated content creation.

Status: Passed both houses of the New York Legislature on June 8, 2026 and awaiting action by Gov. Kathy Hochul. 

Mar. 5, 2025: New York Stop Deepfakes Act (S.6954/A.6540)

Introduced: Mar. 5, 2025 by Sen. Andrew Gounardes; Assembly companion A.6540 sponsored by Assemblymember Alex Bores.

Snapshot: The New York Stop Deepfakes Act would require providers of certain generative AI systems to attach provenance data to synthetic content produced or materially modified by their systems.

Key Provisions: Covered providers would be required to apply provenance data identifying content as synthetic and providing information about its origin and modification, including whether AI was used, the provider involved, and when the provenance data was applied. The legislation would also impose obligations on certain online platforms to preserve provenance data associated with uploaded content and surface that information for certain image, video, and audio content. 

Potential Implications: The proposal would move AI transparency beyond visible labels by embedding information about the origin and modification history of digital content itself. For brands, platforms, publishers, and other businesses producing or distributing digital content, it reflects the growing regulatory push toward standardized content credentials and traceable AI-generated media.

Status: Passed by the Legislature. The Senate passed S.6954-B on June 3, 2026, and the Assembly passed the measure on June 4, 2026. The legislation awaits action by Gov. Kathy Hochul.

Apr. 9, 2025: Content Origin Protection and Integrity from Edited and Deepfaked Media (“COPIED”) Act of 2025 (S.1396)

Introduced: Apr. 9, 2025.

Snapshot: The COPIED Act would establish federal standards and protections for content provenance information, aimed at making it easier to determine the origin and history of digital content and whether it has been generated or modified using AI.

Key Provisions: The legislation would direct the National Institute of Standards and Technology to develop standards and guidelines for content provenance technologies, including methods for identifying the origin and modification history of digital content. It would prohibit knowingly removing, altering, or disabling provenance information in furtherance of unfair or deceptive conduct and impose related obligations on covered platforms. The bill would also restrict certain commercial uses of protected content for AI training or the generation of synthetic content without the copyright owner’s express, informed consent. 

Potential Implications: The proposal sits at the intersection of AI transparency and intellectual property. For brands, publishers, photographers, designers, platforms, and other content owners, standardized provenance tools could provide a way to authenticate original material, identify AI modification, and preserve information about how content may be used.

Status: Pending in the 119th Congress.

Jan. 14, 2025: New York – Publishers of Books Created with Generative AI (S.1815/A.1509)

Introduced: Jan. 14, 2025 in the Senate by Sen. Nathalia Fernandez; Assembly companion A.1509 introduced Jan. 10, 2025.

Snapshot: The legislation would require publishers to disclose when a book sold in New York was created wholly or partially using generative AI.

Key Provisions: Covered printed and digital books would be required to conspicuously disclose the use of generative AI. The proposal applies broadly to books containing text, pictures, audio, puzzles, games, or combinations of those forms. 

Potential Implications: Although directed specifically at publishing, the measure illustrates the growing push toward product-level disclosure of AI-generated creative content and could provide a model for similar requirements affecting other commercial creative works.

Status: Pending in the 2025–26 legislative session. The Senate bill advanced from the Internet and Technology Committee in April 2026 and was committed to the Consumer Protection Committee. 

Sept. 19, 2024: California AI Transparency Act (SB 942; amended by AB 853)

Introduced: SB 942 introduced Jan. 3, 2024 by Sen. Josh Becker and signed into law Sept. 19, 2024; subsequently amended by AB 853, signed Oct. 13, 2025.

Snapshot: California’s AI Transparency Act establishes disclosure and provenance requirements aimed at making AI-generated and AI-modified content identifiable, with AB 853 subsequently expanding and revising the original framework.

Key Provisions: The law requires covered generative AI providers to make tools available that enable users to determine whether specified image, video, and audio content was created or altered by their systems and requires specified disclosures to accompany AI-generated content. AB 853 expanded the framework to address the preservation and display of provenance information by covered platforms.

Potential Implications: The law creates one of the more developed state-level frameworks for identifying and tracing AI-generated content, with implications for AI developers, platforms, advertisers, brands, publishers, and other companies creating or distributing synthetic media.

Status: Enacted; the framework established by SB 942 was subsequently amended by AB 853, which was signed into law Oct. 13, 2025.

AI Agents, Algorithmic Decision-Making, Pricing & Consumer Protection

July 21, 2026: Artificial Intelligence Access, Gatekeeper Exchange, and Nondiscriminatory Transfer Act (AI AGENT Act) (S.5051)

Introduced: July 21, 2026 by Sen. Mark Warner (D-VA).

Snapshot: The AI AGENT Act would require large online platforms to permit qualifying, user-authorized AI agents to access and interact with their services on users’ behalf, subject to registration, privacy, security, and other requirements. The proposal is aimed at promoting competition and reducing switching costs as consumers increasingly use AI agents to interact with online services, including e-commerce platforms.

Key Provisions: The bill would require large online platforms with more than 50 million U.S. customers or subscribers to provide registered “custodial user agents” with access on fair, reasonable, and nondiscriminatory terms and maintain interoperability interfaces for that purpose. AI agents would be required to act within the scope of users’ express authorization, safeguard user data, avoid using that data for advertising, behavioral profiling, or other secondary commercial purposes, maintain records of actions taken on users’ behalf, and comply with duties of care and loyalty.

The Federal Trade Commission would oversee registration and enforcement, while federal agencies would develop standards for interoperability, authentication, and verifiable delegation.

Potential Implications: For retailers, marketplaces, and other large online platforms, the bill could limit their ability to block third-party AI agents from accessing consumer-facing services and require them to build infrastructure that enables agent-led interactions. It would also establish rules governing how AI agents can act for consumers in e-commerce, including around authorization, data use, recordkeeping, and responsibility for agent actions.

Status: Pending. The bill was referred to the Senate Committee on Commerce, Science, and Transportation on July 21, 2026.

May 1, 2026: Colorado – Automated Decision-Making Technology (SB 26-189)

Introduced: May 1, 2026 by Sens. Robert Rodriguez (D) and James Coleman (D), with Reps. Monica Duran (D) and Jennifer Bacon (D).

Snapshot: Colorado’s SB 26-189 repeals and reenacts the state’s earlier AI consumer-protection framework with new requirements governing automated decision-making technology used to make or materially influence consequential decisions.

Key Provisions: The law requires developers of covered automated decision-making technology to provide deployers with specified technical documentation, including information about intended uses, training-data categories, known limitations, and appropriate human review. Deployers must provide consumer notices, including post-adverse-outcome disclosures, and consumers have rights to request certain personal data, correct factually inaccurate data, and seek meaningful human review and reconsideration. Developers and deployers must retain compliance records for at least three years. The Colorado attorney general is responsible for enforcement under the Colorado Consumer Protection Act.

Potential Implications: The law narrows and replaces Colorado’s earlier AI framework while retaining obligations for businesses that use automated systems in consequential areas, including employment, housing, financial and lending services, insurance, education, health care, and certain government services. Companies deploying covered systems will need to assess whether their tools fall within the law and build disclosure, review, data-correction, and recordkeeping processes around qualifying decisions.

Status: Enacted. Gov. Jared Polis signed SB 26-189 on May 14, 2026. The act became law on that date, while its principal requirements for covered automated decision-making technology apply beginning January 1, 2027.

Jan. 21, 2026: New York Artificial Intelligence Civil Rights Act (A.9654)

Introduced: Jan. 21, 2026 by Assemblymember Michaelle Solages.

Snapshot: The New York Artificial Intelligence Civil Rights Act would establish broad protections governing the use of computational algorithms in consequential decisions, including decisions affecting employment, banking, health care, public accommodations, and other areas of economic and civic life.

Key Provisions: The legislation would prohibit developers and deployers from offering, licensing, or using covered algorithms that discriminate based on protected characteristics or produce prohibited disparate impacts. Covered developers and deployers would also be required to conduct independently audited pre-deployment evaluations and post-deployment impact assessments designed to identify and mitigate discriminatory outcomes. The bill includes notice and disclosure requirements, protections for human alternatives, whistleblower protections, and a private right of action. 

Potential Implications: The proposal could impose significant compliance obligations on companies deploying AI and algorithmic systems in consumer- and employee-facing contexts. For retailers and other businesses, its reach could extend to automated systems used in hiring, customer interactions, access to goods and services, and other consequential decisions.

Status: Pending in the New York Assembly.

June 9, 2025: New York Artificial Intelligence Act (S.1169/A.8884)

Introduced: S.1169 introduced Jan. 8, 2025 by Sen. Kristen Gonzalez; Assembly companion A.8884 introduced June 9, 2025 by Assemblymember Michaelle Solages.

Snapshot: The New York Artificial Intelligence Act would establish a regulatory framework for certain high-risk AI systems, with a particular focus on algorithmic discrimination and systems used to make consequential decisions.

Key Provisions: The legislation would prohibit specified discriminatory uses of AI and impose obligations on developers and deployers of high-risk AI systems. Among other things, it calls for independent audits, risk-management programs, reporting requirements, and protections against discriminatory outcomes. It would also prohibit certain social-scoring AI systems. 

Potential Implications: The measure could affect retailers, brands, platforms, and other businesses using high-risk AI systems in areas such as employment and consumer decision-making, requiring companies to evaluate systems for discriminatory impacts and establish formal governance and risk-management processes.

Status: Pending in the 2025–26 legislative session. The Senate bill has been amended multiple times. 

Mar. 28, 2025: New York Preventing Algorithmic Pricing Discrimination Act (S.7033/A.6765)

Introduced: A.6765 introduced March 12, 2025 by Assemblymember Amanda Septimo; S.7033 introduced March 28, 2025 by Sen. Rachel May.

Snapshot: The Preventing Algorithmic Pricing Discrimination Act would regulate the use of consumer data in algorithmically determined prices, building on New York’s existing personalized-pricing disclosure regime.

Key Provisions: The legislation addresses prices set by algorithms using consumer data and establishes disclosure requirements for covered pricing practices. The proposal is aimed at increasing transparency when businesses use information about individual consumers to determine the prices they see. 

Potential Implications: The proposal is directly relevant to retailers and other consumer-facing businesses experimenting with personalized pricing. It reflects growing legislative scrutiny of the use of consumer data to determine individualized prices and could further constrain how companies deploy algorithmic pricing systems.

Status: Pending in the 2025–26 legislative session. 

Dec. 13, 2024: New York Algorithmic Pricing Disclosure Act

Introduced: Originally proposed during the 2023–24 legislative session; signed into law Dec. 13, 2024.

Snapshot: New York’s Algorithmic Pricing Disclosure Act requires businesses to tell consumers when their personal data is being used by an algorithm to set an individualized price.

Key Provisions: Businesses using personalized algorithmic pricing must provide a clear and conspicuous disclosure alongside the price stating: “THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA.” The law defines personal data broadly as information that identifies or could reasonably be linked, directly or indirectly, to a specific consumer or device and provides specified exemptions. 

Potential Implications: The law directly affects retailers and other businesses using consumer-level data to personalize prices and represents a significant move toward regulating so-called surveillance pricing. Rather than prohibiting personalized algorithmic pricing outright, the law requires companies to make the practice visible to consumers at the point where the price is displayed.

Status: Enacted and in effect.

Jan. 8, 2025: New York Artificial Intelligence Consumer Protection Act (S.1962/A.768)

Introduced: A.768 introduced Jan. 8, 2025; Senate companion S.1962 subsequently introduced in the 2025–26 session.

Snapshot: The New York Artificial Intelligence Consumer Protection Act would regulate AI decision systems used to make consequential decisions affecting consumers and prohibit discriminatory uses of those systems.

Key Provisions: The proposal would prohibit covered AI decision systems from discriminating against protected classes and establish requirements concerning bias and governance audits. It defines covered decision systems broadly to include computational processes derived from machine learning, statistical modeling, data analytics, or AI that substantially assist or replace discretionary decision-making in consequential consumer decisions. 

Potential Implications: The measure could require businesses using automated decision systems in consequential consumer contexts to conduct formal bias assessments and evaluate whether their systems produce discriminatory outcomes.

Status: Pending in the 2025–26 legislative session. 

Sept. 21, 2023: Algorithmic Accountability Act of 2023 (S.2892/H.R.5628)

Introduced: Sept. 21, 2023 by Sen. Ron Wyden (D-OR) in the Senate and Rep. Yvette Clarke (D-NY) in the House.

Snapshot: The Algorithmic Accountability Act would have directed the Federal Trade Commission to require certain companies to assess the impacts of automated decision systems used to make or facilitate critical decisions.

Key Provisions: Covered entities would have been required to conduct impact assessments of automated decision systems and augmented critical-decision processes, including evaluations of potential discriminatory, privacy, security, and other harms. The legislation contemplated FTC oversight and reporting requirements intended to create greater accountability around consequential automated systems. 

Potential Implications: The proposal represented an important federal attempt to move algorithmic accountability from voluntary AI-governance practices toward mandatory impact assessments. Its framework anticipated many of the audit, risk-assessment, and documentation requirements that are now appearing in state-level AI legislation.

Status: Introduced in the 118th Congress; did not become law before the end of the congressional session. 

AI Systems, Safety & Governance

Sept. 30, 2026: California – Lawyer Use of Generative AI (SB 574)

Enacted: Sept. 30, 2026; effective Jan. 1, 2027.

Snapshot: California’s SB 574 establishes statutory requirements governing attorneys’ use of generative AI, including rules around professional responsibility, confidential information, verification of AI-generated output, and disclosure of AI use in court submissions.

Key Provisions: The law permits attorneys to use generative AI to assist in the practice of law but prohibits them from delegating the practice of law to AI. Attorneys using generative AI must protect confidential, personally identifying, and other nonpublic information, take reasonable steps to verify AI-generated output and correct erroneous or hallucinated material they use, and disclose the use of generative AI for documents submitted to a court. The law separately requires the attorney responsible for submitting a filing in a California court to personally verify all citations, regardless of whether AI was used. SB 574 also restricts arbitrators from delegating decision-making to generative AI and imposes requirements when they rely on AI-generated information outside the record. Pasted text

Potential Implications: The law creates express statutory guardrails for AI-assisted legal work while preserving attorney responsibility for the resulting work product. For companies and their legal departments, the requirements may affect internal AI policies, approved tools, information controls, litigation workflows, and expectations for outside counsel using generative AI. Pasted text

Status: Enacted as Chapter 858 of the Statutes of 2026 on Sept. 30, 2026; effective Jan. 1, 2027.

Sept. 23, 2026: Ban Artificial Superintelligence Act

Introduced: Sept. 23, 2026 by Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX).

Snapshot: The Ban Artificial Superintelligence Act would prohibit the development and deployment of artificial superintelligence, temporarily pause development of certain advanced AI systems, and create a new federal agency responsible for regulating advanced AI development.

Key Provisions: The legislation would prohibit the development or deployment of AI systems that exceed human cognitive performance across most domains or possess capabilities sufficient to destroy or disempower humanity. It would also pause advanced AI development until a newly created Department of Artificial Intelligence establishes federal safety rules and a model-review process. The department would oversee frontier AI systems, monitor them for specified dangerous capabilities, and enforce the legislation’s restrictions. The proposal also calls for international coordination and export controls aimed at preventing the development of prohibited systems outside the United States.

Potential Implications: The proposal represents a substantially more restrictive approach to federal AI governance than disclosure, transparency, or risk-management legislation. If enacted, it could directly constrain the development and deployment of advanced AI models and create a new federal regulatory regime for companies developing frontier AI systems, with downstream implications for businesses that rely on those models and services.

Status: Pending. Introduced Sept. 23, 2026.

Sept. 9, 2026: California – Independent Verification Organizations (SB 813)

Introduced: 2025 by Sen. Jerry McNerney; signed into law Sept. 9, 2026.

Snapshot: California’s SB 813 establishes a framework for independent organizations that assess the risks posed by AI systems and models, creating a state structure for third-party AI evaluation and verification.

Key Provisions: The law requires the California Government Operations Agency to develop procedures and criteria for designating “independent verification organizations” with demonstrated expertise in assessing risks posed by AI systems or models and identifying appropriate assessment metrics and methodologies. The agency must convene working groups to help identify standards and develop the designation framework, while designated organizations will be subject to reporting requirements.

Potential Implications: The law moves California toward a formal infrastructure for independent evaluation of AI systems, potentially making third-party assessments an increasingly important component of AI governance and compliance. For AI developers, deployers, and companies relying on third-party AI systems, the framework could influence how model risks are evaluated and how compliance with future AI requirements is demonstrated.

Status: Enacted as Chapter 179 of the Statutes of 2026 on Sept. 9, 2026. The Government Operations Agency must take specified steps to establish the independent-verification framework by Jan. 1, 2028.

Sept. 9, 2026: California – AI Auditor Registration (AB 1405)

Introduced: Feb. 21, 2025 by Assemblymember Rebecca Bauer-Kahan; signed into law Sept. 9, 2026.

Snapshot: California’s AB 1405 establishes a state registration and oversight framework for certain third-party auditors that assess AI systems, setting standards around auditor independence, objectivity, integrity, and transparency.

Key Provisions: The law requires the Government Operations Agency to establish an AI Auditor Registry by Jan. 1, 2029 and, beginning on that date, generally prohibits unregistered persons from offering, selling, or conducting covered AI audits. Registered auditors will be required to provide specified information to the state, comply with independence and objectivity requirements, and provide auditees with reports containing specified information about completed audits. The law also establishes whistleblower protections and authorizes the agency to investigate alleged violations and remove auditors from the registry.

Potential Implications: The law creates a formal oversight regime for a growing AI-auditing industry and could affect companies that rely on independent audits to evaluate AI systems or demonstrate compliance with emerging AI requirements. Together with California’s SB 813 framework for independent verification organizations, the measure signals a move toward formalizing the third-party infrastructure used to assess AI systems and their risks.

Status: Enacted as Chapter 178 of the Statutes of 2026 on Sept. 9, 2026. The AI Auditor Registry must be established by Jan. 1, 2029.

California – Companion Chatbot Laws (SB 243 & SB 1119)

Enacted: SB 243 signed into law Oct. 13, 2025; SB 1119 signed into law Sept. 10, 2026.

Snapshot: California has established and subsequently expanded a regulatory framework for AI companion chatbots, imposing disclosure and safety requirements for systems designed to engage users in sustained, human-like social or emotional interactions and establishing additional protections for children who use them.

Key Provisions: SB 243 established baseline requirements for companion-chatbot operators, including disclosure when users could reasonably believe they are interacting with a human, protocols addressing suicide and self-harm, and annual reporting requirements beginning July 1, 2027.

SB 1119 expands the framework for child users. Beginning July 1, 2027, operators must perform and document comprehensive child-safety risk assessments before making new or substantially modified companion chatbots available in California and implement measures to mitigate specified harms. The law also establishes requirements around age assurance, child-safety policies, crisis-response protocols, parental controls, usage limits, advertising and children’s data practices, and independent child-safety audits. It authorizes specified public prosecutors to enforce the requirements and permits children who suffer actual harm from violations, or their parents or guardians acting on their behalf, to bring civil actions against operators. SB 1119 also removes and replaces certain minor-specific provisions of the earlier companion-chatbot framework.

Potential Implications: California’s framework subjects consumer-facing AI products designed to simulate human relationships to increasingly specific product-safety and governance requirements. Companies offering companion chatbots will need to distinguish these systems from ordinary customer-service tools and, where children may use them, build child-safety considerations into product design, risk assessments, age assurance, parental controls, crisis protocols, data practices, and independent auditing.

Status: SB 243 took effect Jan. 1, 2026. SB 1119 was enacted as Chapter 190 of the Statutes of 2026 on Sept. 10, 2026, with its principal child-safety requirements applying beginning July 1, 2027.

March 14, 2025: Texas – Texas Responsible Artificial Intelligence Governance Act (TRAIGA) (HB 149)

Introduced: March 14, 2025 by Rep. Giovanni Capriglione (R-TX).

Snapshot: The Texas Responsible Artificial Intelligence Governance Act establishes a statewide framework regulating specified uses of AI systems, including certain prohibited practices, government and health care disclosure requirements, enforcement mechanisms, and a regulatory sandbox for AI development and testing.

Key Provisions: The law applies to persons that conduct business in Texas, produce products or services used by Texas residents, or develop or deploy AI systems in the state. It prohibits specified uses of AI involving unlawful discrimination, certain biometric-data practices, constitutional rights, social scoring, and certain sexually explicit content. It also imposes disclosure requirements in specified contexts, including certain government-facing consumer interactions and health care uses, creates a regulatory sandbox for AI systems, establishes the Texas Artificial Intelligence Council, and authorizes enforcement by the Texas attorney general, including civil penalties.

Potential Implications: TRAIGA gives companies operating in Texas another state-level AI regime to account for when developing or deploying AI-enabled products and services. Its jurisdictional scope means that technology companies, retailers, platforms, and other businesses offering AI-powered products or services in Texas may need to assess their systems against the law’s prohibited-use, biometric-data, and other compliance provisions, even though the law’s affirmative disclosure obligations are more limited than its overall reach.

Status: Enacted and in effect. Gov. Greg Abbott signed HB 149 on June 22, 2025, and the law took effect January 1, 2026.

Jan. 27, 2025: New York – Advanced AI Licensing Act (A.3356)

Introduced: Jan. 27, 2025 by Assemblymembers Clyde Vanel, Ari Brown, and Alicia Hyndman, with co-sponsors.

Snapshot: The Advanced AI Licensing Act would establish a state licensing and regulatory framework for certain high-risk AI systems.

Key Provisions: The proposal would require covered AI developers or operators to register and comply with specified standards and would authorize the New York Department of State to oversee licensing, suspend or revoke licenses, and impose penalties for violations.

Potential Implications: The measure represents a more interventionist approach to AI governance than disclosure-based legislation, moving toward affirmative licensing of certain high-risk systems. If enacted, it could create direct market-entry and ongoing compliance requirements for companies developing or deploying covered AI technologies in New York.

Status: Pending; referred to the New York Assembly Committee on Science and Technology. 

Nov. 15, 2023: Artificial Intelligence Research, Innovation, and Accountability Act of 2023 (S.3312)

Introduced: Nov. 15, 2023 by Sens. John Thune, Amy Klobuchar, Roger Wicker, John Hickenlooper, Shelley Moore Capito, and Ben Ray Luján.

Snapshot: The Artificial Intelligence Research, Innovation, and Accountability Act proposed a federal framework aimed at increasing transparency, accountability, safety, and security around higher-impact AI systems while preserving room for innovation.

Key Provisions: The legislation contemplated risk-based obligations for certain AI systems, including transparency requirements, testing and evaluation, and additional oversight for higher-impact applications.

Potential Implications: The proposal reflected an early bipartisan federal effort to create differentiated obligations based on AI-system risk rather than regulating all AI uses uniformly. Its approach anticipated the broader move toward risk-based governance that now appears across state and international AI regulation.

Status: Introduced in the 118th Congress; did not become law before the end of the congressional session.