New California bill pushes back against AI copyright infringement

California Assemblymember Rebecca Bauer-Kahan authored last year’s successful Training Data Transparency Act. Now she’s back with a bill to protect performers, authors, and artists from widespread copyright infringement by AI developers.

One of California’s first significant AI bills of the 2025 session dropped earlier this week, and it’s a powerful proposal for AI transparency and intellectual property protection.

Assemblymember Rebecca Bauer-Kahan (D-Orinda), has introduced AB 412, the AI Copyright Transparency Act, aimed at increasing transparency around the use of copyrighted materials to train generative artificial intelligence (GenAI) systems and models.

The proposed legislation requires GenAI developers to inform copyright owners when their materials are included in GenAI training datasets. It also includes one of the first requirements for developers to install a Training Data Verification Request (TDVR) tool, which would allow individuals to query the developer to find out if specific intellectual property was used to train a certain AI model. TCAI has called for the adoption of TDVRs and accompanying Training Data Deletion Request (TDDR) tools as a required industry standard.

AB 412 would require AI developers to include a Training Data Verification Request tool to allow individuals to see if their material was used to train an AI model. Select image for article.

‘increases accountability for ai developers’

"As the AI industry continues to develop and expand, it is critical for content creators to know if and how their work is being used to train advanced models," said Assm. Bauer-Kahan. "The AI Copyright Transparency Act increases accountability for AI developers and empowers copyright owners to exercise their rights."

Transparency Coalition Founder Jai Jaisimha expressed strong support for the proposal.

“We are especially excited to see the incorporation of Training Data Verification Requests,” said Jaisimha, “which we have been advocating for since the passage last year of AB 2013 Training Data Transparency Act authored by Assm. Jacqui Irwin.”

Assm. Bauer-Kahan’s new bill, AB 412, “would provide additional visibility to creators and allow them to understand if their data is being used to train a generative AI model,” Jaisimha added.

“We also will be urging the incorporation of a Training Data Deletion Request tool, similar to California's pioneering Delete Act for personal data, in future versions of this bill and look forward to supporting its passage."

using intellectual property without consent

AI developers often use copyrighted material to train new GenAI models without crediting or compensating the owner of that material.

Dozens of lawsuits have been filed that accuse major AI developers like OpenAI (maker of ChatGPT), Microsoft, Meta, and Anthropic of widespread copyright infringement for using pirated books, movies, recordings, and other content to train their AI models.

Those lawsuits are slowly working their way through federal court in California, New York, and other jurisdictions. TCAI has a curated guide to the most important lawsuits.

Courts move slowly. AI developers move quickly—and are continuing to use copyrighted material to train their AI models and build billion-dollar companies on the backs of uncompensated and unacknowledged writers, performers, and producers.

key elements of the ai copyright transparency act

AB 412 includes these provisions:

• Right to Information: Requires developers to provide copyright owners with lists of copyrighted materials used to train GenAI systems and models.

• Online Request Mechanism: Requires developers to establish a publicly accessible mechanism allowing copyright owners to submit requests for information.

• Enforcement: Authorizes copyright owners to bring civil actions against developers who fail to provide required information.

was your copyrighted material used without consent?

There are two limited search engines available for writers and artists to see if their work is collected in two of the pirated-content datasets most commonly used by big tech corporations. TCAI has a portal to those search engines.

Those are merely two of the many databases out there, however. Copyright owners have no way of truly knowing whether any particular material has been used to train a given model. AB 412 would remedy that by requiring AI developers to inform copyright holders when their material has been used to train a model.

Currently there are only two limited search tools available to authors and artists. Select image for article.

The Copyright Transparency Act is sponsored by the Screen Actors Guild - American Federation of Television and Radio Artists (SAG-AFTRA) and co-sponsored by the Concept Art Association (CAA) and the National Association of Voice Actors (NAVA).

SAG-AFTRA National Executive Director and Chief Negotiator Duncan Crabtree-Ireland said: “This is the type of commonsense legislation we need to see more of in the A.I. age. Individuals and companies deserve to know if their intellectual property has been used to train A.I. models. SAG-AFTRA applauds Assemblymember Rebecca Bauer-Kahan for introducing this bill to protect human artistry and ensure A.I. remains accountable to the work of human content creators.”

The Concept Art Association also issued a statement of support: ”Thousands of illustrators have found the entirety of their works, both personal and professional, used to train Generative Artificial Intelligence (GenAI) datasets without anyone having acquired consent, or having offered them any kind of credit or compensation. Vast quantities of data are routinely and indiscriminately scraped from across the internet to train these GenAI systems and models. Because there is no way for creators to verify if their works have been used, its only logical for them to assume that their copyrighted materials are contained within in these systems. There is a great need for transparency in this realm.”