Politics

California bill aims to curb AI training without creators’ consent amid industry pushback

Thursday, 18 December 2025 6:07AM UTC

Artists, writers, and actors rally in California to support AB 412, a bill promoting transparency and consent in AI training datasets, amid industry opposition and technological debates.

Stanford’s Paul Brest Hall was so packed I had to find a spot to sit on the floor, the lead reporter wrote after a Monday morning hearing on December 8th where artists, writers and actors turned out in force to back Assembly Bill 412, the AI Copyright Transparency Act. According to the lead account in Blood in the Machine, every chair was taken; spectators lined the walls and shuffled around in the back as creatives sought a legislative check on technology companies training large language and image models on copyrighted work without notice or consent. ^[1]^[2]

AB 412, authored by Assemblymember Rebecca Bauer-Kahan, would require developers of generative AI systems to document what goes into their training datasets and to alert copyright holders if their works were included, the bill text shows. The proposal aims to give creators a way to learn whether their work has been used and to request removal, creating a transparency mechanism proponents say is a necessary first step toward consent, credit and compensation. ^[3]

The hearing brought vivid testimony from industry participants and unions. “I don’t know how many of you have come to LA recently,” Danny Lin, president of the Animation Guild, told the committee, warning that “it is bleeding out in front of my very eyes,” the reporter noted. SAG-AFTRA and the Writers Guild of America West have publicly supported AB 412, with WGAW stressing the bill’s role in protecting writers’ intellectual property and SAG-AFTRA announcing sponsorship and backing for transparency measures to safeguard performers. ^[1]^[4]^[5]

Speakers painted a picture of rapidly eroding livelihoods. Actors, voice artists and animators described job losses, speed-ups and precarity as studios and tech companies increasingly substitute AI-generated output for human work. Jason George, a SAG-AFTRA board member, warned that voice actors “are in trouble” within a few years and cited an AI-generated country song as an example of cultural misappropriation, an anecdote amplified in the hearing account. ^[1]

Legal and technical experts at the hearing offered a more mixed assessment of copyright’s ability to solve the problem. UC Berkeley professor Pamela Samuelson told the committee she believes courts may ultimately find training on copyrighted materials to be fair use in the United States, while Stanford’s Mark Lemley argued documentation requirements could hamper innovation. The Electronic Frontier Foundation has publicly opposed AB 412 on the grounds it could unduly burden smaller AI developers, warning the law might “bury small AI companies,” an argument civil liberties advocates have framed as a trade-off between transparency and innovation. ^[1]^[7]

Countering notions that transparency is infeasible, the lead report and witnesses referenced European steps and technical prototypes that illustrate workable approaches. Gerard de Graaf, head of the EU’s San Francisco office, described EU mechanisms that allow creators to check dataset inclusion and request removal, and University of Chicago researchers demonstrated fingerprinting and “poisoning” tools such as Nightshade and Glaze that can identify or protect copyrighted material. Industry founders and smaller AI firms at the hearing indicated similar technical solutions could be implemented, suggesting a practical path to documentation and opt-out features. ^[1]

The political economy around the bill was starkly visible. The lead piece detailed limited industry attendance, of four invited platforms only one sent a representative, and highlighted how large firms use licensing deals and partnerships as political capital against state-level regulation. That dynamic was underscored by contemporaneous developments: industry consolidation and high-profile deals, including reporting that Disney would enter a major partnership with an AI company to exploit valuable IP, and a federal executive order signed to discourage state AI regulation, both of which the lead account framed as part of a broader effort to pre-empt local protections for creative labour. ^[1]

Supporters argue AB 412 is not a cure-all but a modest, necessary intervention. Transparency could force AI developers to be “more intentional” about dataset composition, the lead report argued, potentially nudging companies toward licensing arrangements rather than indiscriminate scraping. Industry advocates and unions alike see documentation as a lever to expose corporate practices and create bargaining power for creators, even as they acknowledge a single database or occasional settlement payments will not reverse the larger automation trend. ^[1]^[4]^[5]^[6]

Opponents caution about unintended consequences. The EFF and some legal scholars warn documentation rules could advantage well‑capitalised incumbents and chill experimentation by smaller teams, while critics of copyright-based remedies argue labour protections, industrial policy and other tools may be better suited to guard employment and working conditions. The hearing, and the broader debate, thus lays bare a fundamental choice about whether to regulate inputs to model training, reshape labour markets, or both. ^[1]^[7]

The hearing’s human drama was unmistakable: a long line of creative workers given 15 seconds each to speak, many with trembling voices, imploring legislators to act. For many in the room the demand was simple, “AI training on copyrighted material without consent is theft, not fair use,” as one commenter put it in the published account, and the crowd’s near-uniform support for AB 412 signalled a growing political mobilisation among creatives. Whether that mobilisation will translate into durable legal protections or broader policy reforms remains uncertain, but the hearing crystallised the immediate stakes: transparency as an opening gambit in a fight over who will be permitted to profit from cultural production. ^[1]^[2]^[4]

As legislators and technologists weigh policy options, the debate is likely to remain contested and multifaceted. Assembly Bill 412 represents one pathway toward greater accountability; industry groups and civil liberties advocates warn of costs and perverse effects; technical researchers and international regulators demonstrate feasible alternatives. The question the hearing left resonant was not whether art will survive AI but who will have the opportunity to make it, and under what terms. “There has to be a way,” the lead reporter concluded, reflecting the room’s mix of fear, resolve and the search for systemic answers. ^[1]^[3]^[6]^[7]

##Reference Map:

^[1] (Blood in the Machine) - Paragraph 1, Paragraph 4, Paragraph 6, Paragraph 7, Paragraph 8, Paragraph 9, Paragraph 10
^[2] (Blood in the Machine summary) - Paragraph 1, Paragraph 9
^[3] (California Assembly bill AB 412 text) - Paragraph 2, Paragraph 10
^[4] (Writers Guild of America West support letter) - Paragraph 3, Paragraph 9
^[5] (SAG-AFTRA announcement) - Paragraph 3, Paragraph 9
^[6] (Transparency Coalition legislative update) - Paragraph 8, Paragraph 10
^[7] (Electronic Frontier Foundation analysis) - Paragraph 5, Paragraph 9

Source: Noah Wire Services

More on this

https://www.bloodinthemachine.com/p/there-has-to-be-a-way - Please view link - unable to able to access data
https://www.bloodinthemachine.com/p/there-has-to-be-a-way - An article detailing a packed hearing at Stanford's Paul Brest Hall on December 8th, where artists, writers, and actors gathered to support Assembly Bill 412, the AI Copyright Transparency Act. The bill, authored by Assemblymember Rebecca Bauer-Kahan, aims to require AI companies to document their training models and notify copyright holders if their work is included. The piece highlights testimonies from industry professionals expressing concerns over AI's impact on creative industries and the need for legislative action to protect intellectual property rights.
https://apcp.assembly.ca.gov/system/files/2025-03/ab-412-bauer-kahan.pdf - The official text of Assembly Bill 412, introduced by Assemblymember Rebecca Bauer-Kahan, which seeks to mandate that developers of generative AI models document and disclose any copyrighted materials used in their training datasets. The bill aims to increase transparency and accountability in AI development, allowing copyright owners to be informed about the use of their works and to request their removal if desired.
https://www.wga.org/uploadedfiles/news_and_events/public_policy/ab_412_wgaw_support.pdf - A letter from the Writers Guild of America West (WGAW) expressing strong support for Assembly Bill 412. The WGAW emphasizes the importance of the bill in increasing transparency around the use of copyrighted materials to train generative AI models, ensuring that creators are informed about how their works are used and can protect their intellectual property rights.
https://www.sagaftra.org/sag-aftra-sponsors-ai-transparency-bill - An announcement from SAG-AFTRA regarding their sponsorship of Assembly Bill 412, the AI Copyright Transparency Act. The article discusses the bill's objectives to protect artists, actors, musicians, and other copyright owners by requiring AI developers to disclose when copyrighted materials are used to train generative AI models, thereby safeguarding creators' rights in the age of artificial intelligence.
https://www.transparencycoalition.ai/news/ai-legislative-update-july-18-2025 - An update from the Transparency Coalition on AI legislation, including Assembly Bill 412. The article provides insights into the progress of the bill, which aims to establish a framework for copyright owners to determine whether their registered works were used to train generative AI models, highlighting the ongoing efforts to balance innovation with the protection of intellectual property rights.
https://www.eff.org/deeplinks/2025/07/california-ab-412-stalls-out-win-innovation-and-fair-use - An article from the Electronic Frontier Foundation discussing the stalling of Assembly Bill 412 in California. The piece analyzes the implications of the bill's delay, emphasizing the importance of fair use and innovation in the development of generative AI technologies, and the potential impact on small developers and the open web.

Noah Fact Check Pro

The draft above was created using the information available at the time the story first emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed below. The results are intended to help you assess the credibility of the piece and highlight any areas that may warrant further investigation.

Freshness check

Score: 8

Notes: The narrative reports on a December 8, 2025, hearing in support of Assembly Bill 412, the AI Copyright Transparency Act. The earliest known publication date of similar content is March 18, 2025, when the bill was first introduced. The narrative includes updated data and quotes from the December hearing, justifying a higher freshness score. However, the core content is based on earlier reports, which may indicate recycled material. The narrative is based on a press release, which typically warrants a high freshness score. No discrepancies in figures, dates, or quotes were found. No earlier versions show different figures, dates, or quotes. The narrative includes updated data but recycles older material, which may justify a higher freshness score but should still be flagged.

Quotes check

Score: 9

Notes: The narrative includes direct quotes from the December 8, 2025, hearing. No identical quotes appear in earlier material, indicating potentially original or exclusive content. No variations in quote wording were found.

Source reliability

Score: 7

Notes: The narrative originates from 'Blood in the Machine,' a single-outlet publication. While it provides detailed coverage, the lack of multiple reputable sources raises some uncertainty. The report mentions organizations like SAG-AFTRA and the Writers Guild of America West, which are reputable, but their direct involvement in the report is not verified.

Plausability check

Score: 8

Notes: The narrative's claims about the December 8, 2025, hearing and the support for Assembly Bill 412 are plausible and align with known legislative activities. The report lacks specific factual anchors, such as names, institutions, and dates, which reduces the score and flags it as potentially synthetic. The language and tone are consistent with the region and topic. The structure includes excessive or off-topic detail unrelated to the claim, which may be a distraction tactic. The tone is unusually dramatic, vague, and doesn't resemble typical corporate or official language, which flags it for further scrutiny.

Overall assessment

Verdict (FAIL, OPEN, PASS): OPEN

Confidence (LOW, MEDIUM, HIGH): MEDIUM

Summary: The narrative provides detailed coverage of the December 8, 2025, hearing in support of Assembly Bill 412, incorporating direct quotes and updated data. However, it relies on a single-source publication, lacks multiple reputable sources, and includes excessive or off-topic details, which may be a distraction tactic. The tone is unusually dramatic and vague, not resembling typical corporate or official language, warranting further scrutiny.

AI regulation
copyright
creators' rights