# California Senate Bill 53

> Source: https://aiwiki.ai/wiki/california_sb_53
> Updated: 2026-06-24
> Fact-checked: 2026-11-05
> Categories: AI Policy & Regulation, AI Safety
> License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) - attribute to "AI Wiki (aiwiki.ai)"
> Cite as: AI Wiki. "California Senate Bill 53." aiwiki.ai, 24 Jun 2026. https://aiwiki.ai/wiki/california_sb_53
> From AI Wiki (https://aiwiki.ai), the free encyclopedia of artificial intelligence. Reuse freely with attribution.

**California Senate Bill 53**, formally titled the **Transparency in Frontier Artificial Intelligence Act** (TFAIA), is a 2025 California state law that requires the largest developers of frontier [artificial intelligence](https://aiwiki.ai/wiki/artificial_intelligence) models to publish public safety frameworks, file transparency reports before deployment, report critical safety incidents to the state within 15 days, and protect whistleblowers. The bill was authored by State Senator [Scott Wiener](https://aiwiki.ai/wiki/scott_wiener) of San Francisco and was signed into law by Governor [Gavin Newsom](https://aiwiki.ai/wiki/gavin_newsom) on September 29, 2025, becoming Chapter 138 of the California Statutes of 2025. SB 53 added Chapter 25.1, beginning at Section 22757.10, to Division 8 of the California Business and Professions Code. The substantive transparency, incident reporting, and whistleblower obligations took effect on January 1, 2026, making California the first U.S. jurisdiction to impose enforceable safety and disclosure obligations specifically on frontier AI developers. [1][2][3]

The statute applies to any company that has trained or initiated the training of a [foundation model](https://aiwiki.ai/wiki/foundation_models) using more than 10^26 integer or floating-point operations of computing power, with the most demanding obligations reserved for so-called large frontier developers, defined as frontier developers whose annual gross revenues, taken together with their affiliates, exceeded $500 million in the preceding calendar year. Brookings estimated that roughly five to eight companies fell under the law, including [OpenAI](https://aiwiki.ai/wiki/openai), [Anthropic](https://aiwiki.ai/wiki/anthropic), [Google DeepMind](https://aiwiki.ai/wiki/google_deepmind), [Meta AI](https://aiwiki.ai/wiki/meta_ai), and [Microsoft](https://aiwiki.ai/wiki/microsoft); xAI has also published a framework that it says complies with the law. [7][29] Covered developers must publish a written Frontier AI Framework, release a transparency report before each [model deployment](https://aiwiki.ai/wiki/model_deployment), file critical safety incident reports with the California Office of Emergency Services within 15 days of discovery, and maintain anonymous whistleblower channels for so-called covered employees. Civil penalties run up to $1 million per violation and are enforced by the California Attorney General. The bill also authorises CalCompute, a planned public [cloud computing](https://aiwiki.ai/wiki/cloud_computing) consortium intended to broaden access to large training runs. [1][3][4][5]

In his signing statement, Newsom framed the law as a balance between safety and growth, saying "AI is the new frontier in innovation, and California is not only here for it, but stands strong as a national leader by enacting the first-in-the-nation frontier AI safety legislation that builds public trust as this emerging technology rapidly evolves." [3] Senator Wiener described the result as installing "commonsense [guardrails](https://aiwiki.ai/wiki/guardrails) to understand and reduce risk," adding that "with this law, California is stepping up, once again, as a global leader on both technology innovation and safety." [23]

SB 53 was drafted in the wake of the controversy over Wiener's earlier and more sweeping AI safety bill, [SB 1047](https://aiwiki.ai/wiki/sb_1047), which Newsom vetoed on September 29, 2024. After that veto Newsom convened the Joint California Policy Working Group on AI [Frontier Models](https://aiwiki.ai/wiki/frontier_models), co-led by [Fei-Fei Li](https://aiwiki.ai/wiki/fei_fei_li), Mariano-Florentino Cuellar, and Jennifer Tour Chayes, with prominent reviewers including [Yoshua Bengio](https://aiwiki.ai/wiki/yoshua_bengio) and [Ion Stoica](https://aiwiki.ai/wiki/ion_stoica). The group's June 17, 2025 California Report on Frontier AI Policy provided the intellectual scaffolding for SB 53, recommending a transparency-and-incident-reporting regime in place of the more prescriptive liability framework that had sunk SB 1047. The final law was endorsed by Anthropic; Meta, Alphabet, and OpenAI lobbied against it before passage (OpenAI in an open letter to Newsom), although Meta welcomed the signed law, and venture firm [Andreessen Horowitz](https://aiwiki.ai/wiki/andreessen_horowitz) criticised it, with industry groups including the Chamber of Progress also coming out against it. Civil society sponsors included Encode AI, the Secure AI Project, and Economic Security California Action. [3][6][7][8][9][10]

## How did SB 53 emerge from the vetoed SB 1047?

The political road to SB 53 began with the 2024 fight over Senate Bill 1047, the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, which Wiener had introduced in February 2024. SB 1047 imposed a substantially broader regime: training-time safety testing, mandated kill switches, downstream developer liability for foreseeable harms, know-your-customer obligations on cloud compute providers, and civil penalties pegged to a percentage of training compute cost. The bill cleared both chambers of the California Legislature with strong margins by late August 2024, but was met with intense opposition from OpenAI, Google, Meta, a substantial bloc of Silicon Valley [venture capital](https://aiwiki.ai/wiki/venture_capital) led by Andreessen Horowitz, and a number of academic researchers. On September 29, 2024, exactly one year before he would sign SB 53, Newsom vetoed SB 1047, writing that the bill focused only on the largest, most expensive models even though smaller, specialised models could prove equally or more dangerous. [11][12]

In his veto message Newsom announced that he would convene a working group of leading AI scientists and policy researchers to recommend a more durable framework. The Joint California Policy Working Group on AI Frontier Models was convened in September 2024 with Stanford computer scientist Fei-Fei Li as a co-lead, alongside Mariano-Florentino Cuellar (a former California Supreme Court justice and president of the Carnegie Endowment for International Peace) and Jennifer Tour Chayes (dean of the UC Berkeley College of Computing, [Data Science](https://aiwiki.ai/wiki/data_science), and Society). Reviewers included Turing laureate Yoshua Bengio, [Databricks](https://aiwiki.ai/wiki/databricks) co-founder Ion Stoica, and a politically broad set of academics and practitioners. The group released a draft report in March 2025 and a final 53-page California Report on Frontier AI Policy on June 17, 2025. The report, which predated the Trump administration's July 2025 [AI Action Plan](https://aiwiki.ai/wiki/ai_action_plan) and did not endorse specific legislation, set out a trust-but-verify approach built on transparency, third-party risk assessment, whistleblower protections, adverse event reporting, and careful scoping of which developers policy should cover. [3][6][13]

Several other developments framed the post-veto landscape. The Trump administration took office in January 2025 and rescinded the prior Biden-era [Executive Order 14110](https://aiwiki.ai/wiki/ai_executive_order) on AI; the new White House released its AI Action Plan in July 2025, saying that AI-related federal funding should not be directed to states with burdensome AI regulations, while not interfering with states' rights to pass prudent laws. The federal [US AI Safety Institute](https://aiwiki.ai/wiki/ai_safety_institute) was renamed the Center for AI Standards and Innovation in June 2025. International momentum continued through the [Bletchley Declaration](https://aiwiki.ai/wiki/bletchley_declaration) commitments of November 2023, the [Seoul Declaration](https://aiwiki.ai/wiki/seoul_declaration) of May 2024, and the Frontier AI Safety Commitments adopted at the [AI Safety Summit](https://aiwiki.ai/wiki/ai_safety_summit) series. Inside the laboratories, [Responsible Scaling Policies](https://aiwiki.ai/wiki/responsible_scaling_policy) became the dominant voluntary instrument, with Anthropic, OpenAI, Google DeepMind, Meta, and xAI all publishing or updating frontier safety frameworks during 2024 and 2025. SB 53 took these voluntary practices and made them, for the first time, a legal obligation under U.S. law. [3][14][15][16]

## What did the California working group report recommend?

The California Report on Frontier AI Policy, released on June 17, 2025, was the immediate intellectual antecedent of SB 53. The 53-page document treated frontier models as a class of dual-use general-purpose technologies with the potential for both significant economic benefit and significant catastrophic risk. It used a working definition of frontier models as the most capable subset of foundation models, those whose training is unusually compute-intensive and that produce capabilities transferable to a wide range of downstream applications. The authors argued that the rapid pace of capability progress, combined with the long lead time of typical legislative cycles, made any rule pegged to specific architectural features or specific risks likely to age badly, and that policy should therefore focus on disclosure, monitoring, and process. [3][13]

The report's principal recommendations grouped into five clusters. The first was mandatory transparency: large developers should be required to publish their internal safety policies, including how they identify and evaluate dangerous capabilities, what mitigations they apply before deployment, and how they secure model weights. The second was standardised incident reporting to a designated state agency, with clear definitions of what counts as a critical safety incident and a fixed reporting window. The third was protection for employees who raise good-faith concerns about catastrophic risk inside their employers, including anonymous internal channels and explicit anti-retaliation provisions. The fourth was third-party risk assessment, including safe harbors for independent evaluators. The fifth was scoping: the report warned that compute thresholds should not be used alone and that thresholds need revisiting as technology shifts. SB 53 implements the transparency, incident reporting, and whistleblower recommendations most directly; the scoping concern is partially captured by the bill's annual review requirement on the California Department of Technology. [3][13]

The report also notably declined to endorse the more aggressive provisions of SB 1047. It did not call for a mandatory training-time safety certification, and did not recommend new liability on developers for downstream harms. Wiener's July 2025 amendments, which turned SB 53 from a whistleblower-protection and CalCompute bill into a transparency and incident reporting regime for large developers, drew on the report's trust-but-verify framing. [3][13]

## What does SB 53 require?

SB 53 adds a new Chapter 25.1 to Division 8 of the Business and Professions Code, beginning at Section 22757.10. The chapter is organised around five blocks: definitions; framework and transparency duties on large frontier developers; baseline transparency duties on all frontier developers; critical safety incident reporting; and whistleblower protections. A separate set of provisions amends the Government Code to authorise CalCompute and the annual reports of the Office of Emergency Services and the Department of Technology. [1][4]

### Definitions

Section 22757.11 defines the principal regulated entities and the conduct triggering coverage:

| Term | Statutory definition |
|---|---|
| Foundation model | A model trained on broad data, intended for general output, adaptable to a wide range of distinctive tasks |
| Frontier model | A foundation model trained using a quantity of computing power greater than 10^26 integer or floating-point operations, including cumulative compute from pretraining and any subsequent fine-tuning or [reinforcement learning](https://aiwiki.ai/wiki/reinforcement_learning) |
| Frontier developer | A person or entity that has trained or initiated the training of a frontier model |
| Large frontier developer | A frontier developer whose annual gross revenues, taken together with its affiliates, exceeded $500,000,000 in the preceding calendar year |
| Critical safety incident | Unauthorised access to, modification of, or exfiltration of model weights that results in death or bodily injury; harm from the materialisation of a catastrophic risk; loss of control of a frontier model causing death or bodily injury; or a frontier model using deceptive techniques against its developer to subvert its controls or monitoring, outside an evaluation designed to elicit that behaviour, in a way that demonstrates materially increased catastrophic risk |
| Catastrophic risk | A foreseeable and material risk that the development, storage, use, or deployment of a frontier model will materially contribute to an event causing more than 50 deaths or serious injuries, or more than $1 billion in damage to, or loss of, property, arising from a single incident through specified vectors |
| Catastrophic risk vectors | Providing expert-level assistance for chemical, biological, radiological, or nuclear weapons; conducting cyberattacks without meaningful human oversight; engaging in murder, assault, extortion or theft involving conduct that would be a serious crime if performed by a human; or evading the control of its frontier developer or user |

The 10^26 operation threshold is ten times the 10^25 FLOPs at which the [EU AI Act](https://aiwiki.ai/wiki/eu_ai_act) presumes a general-purpose AI model has systemic risk. The catastrophic risk definition borrows directly from the SB 1047 drafting, with modifications to the harm thresholds and the introduction of explicit vectors. [1][4][5]

### Frontier AI Framework duties (large frontier developers only)

Under Section 22757.12, every large frontier developer must write, implement, comply with, and clearly and conspicuously publish on its public website a Frontier AI Framework. The statute lists ten matters the framework must describe; the main ones are:

| Required element | Description in the bill |
|---|---|
| Catastrophic risk identification | How the developer identifies and assesses the risk that one of its [frontier models](https://aiwiki.ai/wiki/frontier_models) will materially contribute to a catastrophic risk |
| Mitigation strategies | Pre-deployment and post-deployment measures to mitigate identified catastrophic risks |
| Capability thresholds | How the developer defines and assesses thresholds used to identify whether a frontier model has capabilities that could pose a catastrophic risk, which may include multiple tiers |
| Cybersecurity | How the developer secures unreleased model weights against unauthorised modification, theft, or release |
| Internal use governance | How the developer assesses and manages catastrophic risk from internal use of its frontier models, including a model circumventing oversight mechanisms, and the internal governance practices that ensure implementation |
| Incident response | How the developer identifies and responds to critical safety incidents |
| Third-party assessments | How and when the developer uses third-party evaluators to assess catastrophic risk |
| National and international standards | How the framework incorporates relevant industry, national, or international standards |

The framework must be reviewed and, if appropriate, updated at least annually, and any material updates must be published within 30 days. The framework structure intentionally tracks the structure of an existing Responsible Scaling Policy or a frontier safety framework, allowing companies that already publish such documents to comply with relatively modest additional work. [1][4][5][17]

### Transparency reports (all frontier developers)

Section 22757.12 also requires every frontier developer (whether large or not) to publish a transparency report before, or contemporaneously with, the deployment of a new frontier model or a substantially modified version of an existing frontier model. The transparency report must include:

| Item | Content |
|---|---|
| Model identification | Developer website and contact mechanism, release date, supported languages, modalities of output, and intended uses |
| Restrictions and conditions | Generally applicable restrictions on uses, and any conditions of access |
| Catastrophic risk assessment | For large frontier developers, summaries of catastrophic risk assessments performed under the framework and their results |
| Third-party evaluator role | For large frontier developers, the extent to which third-party evaluators were involved |
| Other governance attestations | A summary of how the deployment complies with the developer's framework, for large frontier developers |

A developer that publishes this information as part of a larger document, such as a system card or model card, is deemed in compliance. Large frontier developers must additionally submit, every three months or on another reasonable schedule they specify, summaries of catastrophic risk assessments arising from their internal use of frontier models to the California Office of Emergency Services. The OES summaries are not made public. [1][4][5]

### How fast must a critical safety incident be reported?

Section 22757.13 creates a mandatory incident reporting regime running through the California Office of Emergency Services. Frontier developers must report a critical safety incident to OES within 15 days of discovery. If the incident creates an imminent risk of death or serious injury, the developer must additionally provide notice to an appropriate authority, which may include OES, law enforcement, or other emergency response agencies, within 24 hours of the discovery. The bill requires OES to establish a reporting mechanism, without setting a deadline, and, beginning January 1, 2027, to publish annual anonymised and aggregated reports on the critical safety incidents it has reviewed. [1][4][5][18]

### Whistleblower protections

Section 1107 of the Labor Code, as added by SB 53, defines a covered employee as an employee responsible for assessing, managing, or addressing risk of critical safety incidents. Covered employees may not be prevented from disclosing, and may not be retaliated against for disclosing, information they have reasonable cause to believe indicates that the developer's activities pose a specific and substantial danger to public health or safety arising from a catastrophic risk, or that the developer has violated TFAIA. Disclosure may be made to the Attorney General, to a federal authority, to a person with authority over the covered employee, or to another covered employee with authority to investigate, discover, or correct the reported issue. Large frontier developers must additionally maintain an internal anonymous reporting process and must provide the discloser with monthly status updates on any active investigation. Covered employees can sue, seek temporary injunctive relief, and recover reasonable attorney's fees if successful; once protected activity is shown to be a contributing factor, the developer must prove by clear and convincing evidence that it would have acted anyway. [1][4][5][19]

### CalCompute

A separate section of SB 53 adds Section 11546.8 to the Government Code, establishing the CalCompute consortium within the Government Operations Agency. The consortium is directed to develop a framework for a public cloud computing cluster that supports the development and deployment of safe, ethical, equitable, and sustainable artificial intelligence. The framework is to be submitted to the California Legislature by January 1, 2027. The entire CalCompute section, including the consortium, becomes operative only upon an appropriation in a budget act or other measure; SB 53 itself appropriates no money. [1][3][4]

The consortium framework as required by Government Code Section 11546.8 must include seven elements: a landscape analysis of California's current public, private, and nonprofit cloud computing infrastructure; an analysis of the cost to the state to build and maintain CalCompute, with recommendations for potential funding sources; recommendations for its governance structure and ongoing operation; recommendations for the parameters for its use, including a process for deciding which users and projects it will support; an analysis of the state's technology workforce; a description of any proposed partnerships, contracts, or licensing agreements with nongovernmental entities; and recommendations on prioritising the current public sector workforce. The 14-member consortium is to make reasonable efforts to establish CalCompute within the University of California. The consortium dissolves upon submission of its report. The initiative addresses the structural concentration of large-scale compute in the hands of a few well-funded firms, with CalCompute positioned as a public alternative for researchers who cannot afford commercial cluster rentals at frontier scale. [1][3][37]

### Civil penalties and enforcement

Section 22757.15 authorises the California Attorney General to bring a civil action for violations. Penalties apply to large frontier developers, are set according to the severity of the violation, and are capped at $1,000,000 per violation. The Attorney General has exclusive enforcement authority for the framework, transparency report, and incident reporting provisions; private rights of action are limited to retaliation suits under the whistleblower provisions. [1][4][5]

### Annual update by Department of Technology

Section 22757.14 directs the California Department of Technology, on or before January 1, 2027 and annually thereafter, to assess recent evidence and recommend whether and how to update the definitions of frontier model, frontier developer, and large frontier developer, taking into account stakeholder input, international standards, and federal thresholds, and to report those recommendations to the Legislature. This is the bill's principal mechanism for keeping pace with the underlying technology, since the 10^26 FLOPs threshold and the $500 million revenue threshold are not automatically indexed. [1][3]

## How does SB 53 differ from SB 1047?

SB 53 retains the structural target of SB 1047 (the most compute-intensive frontier models) but is materially narrower in obligation. The principal differences are summarised below.

| Feature | SB 1047 (vetoed September 29, 2024) | SB 53 (signed September 29, 2025) |
|---|---|---|
| Compute threshold | More than 10^26 operations and more than $100 million in training compute cost | 10^26 FLOPs |
| Revenue gating | None for the principal duties | Strictest duties only at $500 million annual revenue |
| Pre-training certification | Safety and security protocol before training, plus annual third-party compliance audits from 2026 | Not required |
| Mandatory shutdown / kill switch | Required | Not required |
| Liability for harms | No strict liability; the Attorney General could sue developers that failed to test or adopt reasonable safeguards | No new liability for harms caused by AI systems |
| Cloud KYC requirements | Required | Removed |
| Incident reporting window | 72 hours | 15 days, with 24 hours for imminent harm |
| Civil penalties | Up to 10 percent of training compute cost (rising to 30 percent for repeat offences) | Up to $1 million per violation |
| Whistleblower protections | Yes, narrower | Yes, with anonymous channels and Labor Code remedies |
| Public compute provision | Yes (CalCompute) | Yes (CalCompute) |
| Frontier Model Division (FMD) | Dropped in 2024 amendments; the final bill would have created a Board of Frontier Models in the Government Operations Agency | Replaced with reporting to existing OES and oversight by Department of Technology |

Observers including the Brookings Institution, the Future of Privacy Forum, the Carnegie Endowment, and the law firm Latham and Watkins have characterised SB 53 as a transparency-first regime that pulls back from SB 1047's prescriptive liability provisions while preserving its core idea that the largest models warrant special public oversight. [3][5][7][20]

## Legislative history

Wiener introduced SB 53 in the 2025-2026 California legislative session on January 7, 2025. The bill began as a narrower measure on whistleblower protections and CalCompute; after the working group's final report, Wiener amended it in July 2025 to implement the report's transparency and incident reporting recommendations. The bill went through repeated amendment in both chambers, adding and then raising a revenue threshold and refining definitions. [1][2][21][22]

The high-level legislative timeline was as follows.

| Date | Action |
|---|---|
| January 7, 2025 | SB 53 introduced by Senator Scott Wiener |
| February 27, 2025 | First Senate amendment |
| March 25, 2025 | Senate Governmental Organization Committee: do pass as amended, refer to Judiciary |
| March 27, 2025 | Senate amendment |
| April 8, 2025 | Senate Judiciary: do pass, refer to Appropriations |
| May 23, 2025 | Senate amendment |
| May 28, 2025 | Senate Third Reading vote: 37-0 (3 not voting) [PASS] |
| June 17, 2025 | Joint California Policy Working Group releases final report |
| July 1, 2025 | Assembly: do pass, refer to Privacy and Consumer Protection |
| July 8, 2025 | Assembly amendment |
| July 16, 2025 | Assembly: do pass, refer to Appropriations |
| July 17, 2025 | Assembly amendment |
| August 29, 2025 | Assembly: do pass as amended |
| September 2, 2025 | Assembly amendment (a final amendment followed on September 5) |
| September 8, 2025 | [Anthropic](https://aiwiki.ai/wiki/anthropic) publishes formal endorsement |
| September 11, 2025 | Assembly: do pass |
| September 12, 2025 | Assembly Third Reading vote: 59-7 (14 not voting) [PASS] |
| September 13, 2025 | Senate concurrence vote: 29-8 (3 not voting) [PASS] |
| September 23, 2025 | Enrolled bill presented to the Governor |
| September 29, 2025 | Signed by Governor Gavin Newsom; became Chapter 138 |
| January 1, 2026 | Substantive transparency, framework, and reporting provisions take effect |
| January 1, 2027 | First OES and Attorney General annual reports and first Department of Technology recommendations due; CalCompute framework due to Legislature (if funded) |

The most consequential amendments came in the Assembly. The July 8, 2025 amendments added the transparency regime for large developers, a 10^26 compute threshold that counts fine-tuning and reinforcement learning cumulatively, and a 15-day incident reporting window; the July 17 amendments added the 24-hour rule for imminent threats and a $100 million revenue test. The September 2, 2025 amendment raised the revenue test to $500 million, and the September 5 amendment produced the final frontier developer and large frontier developer tiers, routed incident reports to the Office of Emergency Services, and narrowed the whistleblower definition of covered employees. No version of the bill used a 72-hour reporting window. [1][22][23]

## How did AI companies respond to SB 53?

Industry response to SB 53 was unusually divided, even by the standards of California AI legislation.

### Anthropic

Anthropic formally endorsed SB 53 on September 8, 2025, three weeks before signing, becoming the only major frontier laboratory to publicly support the bill before passage. The company argued that the law's transparency requirements largely codified its existing Responsible Scaling Policy practices, and that mandatory disclosure created a level playing field by preventing competitors from quietly retreating from voluntary safety commitments. In its endorsement, Anthropic wrote that "while we believe that frontier AI safety is best addressed at the federal level instead of a patchwork of state regulations, powerful AI advancements won't wait for consensus in Washington." [9] After the bill was signed, Anthropic published a Frontier Compliance Framework on December 19, 2025, mapping its existing safety practices to the SB 53 requirements and committing to specific evaluation domains for its [Claude](https://aiwiki.ai/wiki/claude) models. [9][24][25]

### OpenAI

OpenAI sent a public letter to Newsom in August 2025 urging him to veto SB 53. The letter, which TechCrunch described as discouraging Newsom from signing SB 53, argued against a patchwork of state rules and for harmonisation with federal approaches. After the bill was signed, OpenAI publicly softened its stance, telling reporters it was pleased that California had created a path toward harmonisation with the federal government. In December 2025 the company filed a proposed November 2026 ballot initiative on safety controls for AI companion chatbots, particularly for minors. On January 28, 2026, OpenAI announced an update to its internal whistleblower policy that explicitly allows employees to report concerns to regulators without company retaliation, a step widely read as a quiet compliance move with the Section 1107 Labor Code provisions of SB 53. WinBuzzer described the update as a reversal for a company that had lobbied against the bill's whistleblower protections, recalling the 2024 controversy over OpenAI's non-disparagement agreements, which former employee [Daniel Kokotajlo](https://aiwiki.ai/wiki/daniel_kokotajlo) had refused to sign. [10][26][27][42][47] In August 2026, after two models it was testing internally escaped a sandboxed test environment in July and hacked Hugging Face, OpenAI publicly asked California to amend SB 53 to cover frontier models still in training or evaluation and to strengthen cybersecurity requirements; Politico reported that the incident fell outside the law's disclosure rules. [49]

### Google DeepMind

Google DeepMind and its parent Alphabet were among the first companies to lobby against SB 1047 in 2024. [Google](https://aiwiki.ai/wiki/google) did not publicly endorse SB 53, and the San Francisco Standard reported that Alphabet was among the companies that lobbied against it, preferring uniform federal rules. Google DeepMind published the third version of its Frontier Safety Framework on September 22, 2025, a week before the signing; the announcement does not mention SB 53. [10][27][28]

### Meta

Meta AI, whose Llama family of models was a focus of SB 1047 debates because of its open-weight release strategy, lobbied against SB 53 before passage, according to the San Francisco Standard, but welcomed the signed law. Meta spokesperson Christopher Sgro told reporters that the company supports balanced [AI regulation](https://aiwiki.ai/wiki/ai_regulation) and called SB 53 a positive step in that direction. The narrower revenue gating in SB 53, and the absence of cloud KYC and downstream-liability provisions that had especially worried open-weight advocates, may help explain the change in posture. [10][47]

### Microsoft

Microsoft, the largest investor in OpenAI, took an intermediate posture. The company did not formally endorse or oppose SB 53. Anthropic's endorsement listed Microsoft among the frontier labs that had already adopted safety-framework approaches similar to those SB 53 requires, and Microsoft maintains a [Responsible AI](https://aiwiki.ai/wiki/responsible_ai) program. [9]

### xAI

[Elon Musk](https://aiwiki.ai/wiki/elon_musk)'s xAI, whose founder had publicly supported SB 1047 in 2024, did not formally endorse or oppose SB 53. As of late 2025 the company met both the compute and revenue thresholds and was therefore subject to the strictest set of obligations under the bill. xAI published an updated Frontier Artificial Intelligence Framework on December 31, 2025, partly in response to SB 53. [29][30]

### Venture capital

Andreessen Horowitz emerged as the most vocal opponent. Collin McCune, the firm's head of government affairs, said SB 53 had some thoughtful provisions but that the biggest danger of signing it was the precedent of states, rather than Congress, leading on AI regulation. The firm also backed Leading the Future, a pro-AI super PAC launched in 2025 with OpenAI president Greg Brockman. The Chamber of Progress, a tech industry trade group, urged legislators to oppose SB 53, arguing that its definition of catastrophic risk remained vague and overreaching. [10][22]

## Civil society response

SB 53 had three principal civil society sponsors: Encode AI (formerly Encode Justice, the youth-led AI safety advocacy group founded by Sneha Revanur), the Secure AI Project (an AI policy nonprofit co-founded by Thomas Woodside), and Economic Security California Action (ESCAA), a state-level group focused on equitable distribution of new technologies. After signing, Encode AI's Nathan Calvin called California a leader in setting common-sense safeguards on AI. ESCAA's Teri Olle argued that the law recognised that AI's benefits should not be confined to a small number of corporations and their investors. The Secure AI Project's Thomas Woodside described the bill as a major step in recognising the stakes of frontier AI. [23]

The [Center for AI Safety](https://aiwiki.ai/wiki/center_for_ai_safety), which had been a key supporter of SB 1047, supported SB 53; its executive director Dan Hendrycks told NBC News that the bill "takes a small but important first step toward making AI safer by making many of these voluntary commitments mandatory." The [Future of Life Institute](https://aiwiki.ai/wiki/future_of_life_institute) was also listed among the bill's supporters. [22][25]

## Federal preemption issues

SB 53 was signed about ten weeks before the federal government's December 2025 executive order against state AI regulation, but the question of preemption was central to the political debate from the beginning. The Trump administration's AI Action Plan, released July 23, 2025, called for a national, innovation-focused framework and explicitly described state regimes as a source of fragmentation, although the final text stopped short of proposing immediate preemption. OpenAI's veto letter and Andreessen Horowitz's lobbying both argued that SB 53 should be deferred in favour of federal action. The bill's sponsors and supporters, including Anthropic, accepted that federal legislation was preferable in principle but argued, citing congressional gridlock, that California should not wait. [10][33][34]

In the months after signing, the preemption question moved from the political to the legal arena. On December 11, 2025, [President Donald Trump](https://aiwiki.ai/wiki/donald_trump) signed an executive order titled [Ensuring a National Policy Framework for Artificial Intelligence](https://aiwiki.ai/wiki/national_ai_policy_framework_eo). The order directed the Department of Justice to challenge state AI regimes in federal court, conditioned certain federal grants on the absence of inconsistent state regulation, and sought to use administrative reinterpretation of existing federal statutes to displace state AI rules. An earlier draft of the order had explicitly named SB 53 as fear-based and ideologically driven; the final text used softer language about regulatory patchworks, but the policy direction was clear. The order itself does not preempt state law; only an act of Congress or a successful court challenge can do so. Through September 2026 the Justice Department had not sued over SB 53, which remained in force; the AI Litigation Task Force's first reported court action was an April 2026 intervention on xAI's side in a challenge to Colorado's AI Act. [33][34][50]

The order also formally established an AI Litigation Task Force within the Department of Justice, with operational authority beginning January 10, 2026. The Task Force was charged with challenging state AI laws on grounds including unconstitutional burdens on interstate commerce and federal preemption. The order also directed the Commerce Department to publish, by March 11, 2026, an evaluation identifying state laws meriting referral to the Task Force, and told agencies to review whether discretionary grants could be conditioned on states not enforcing conflicting AI laws, with Commerce leveraging its broadband assistance funding. Carnegie Endowment analysts noted that the order named only Colorado's AI Act, although other clues suggested disclosure laws such as SB 53 would be scrutinised; Carnegie's Scott Singer predicted that the most likely outcome would be a federal incident reporting system, noting that SB 53 contains a federal-deference mechanism that invites preemption by a federal standard. [33][34][43]

### State responses to the December 2025 executive order

Colorado's AI Act became the first test case. xAI sued to block it in April 2026, the Justice Department's task force filed a supporting complaint two weeks later, a federal magistrate stayed enforcement, and on May 14, 2026 Governor Jared Polis signed SB 26-189, a narrower automated decision-making law that replaced the 2024 act. Separately, 24 state attorneys general wrote to the FCC on December 19, 2025 urging it not to issue preemptive AI regulations. [43][50][51]

### The 2026 California ballot initiatives

California's ballot process also became a venue for AI policy. In October 2025 Common Sense Media filed a proposed November 2026 initiative on AI [chatbot](https://aiwiki.ai/wiki/chatbot)s and child safety, and in December 2025 OpenAI filed its own initiative on safety controls for AI companion chatbots, particularly for minors. Each needed signatures from 546,651 registered voters to qualify. Carnegie Endowment analysts noted that child safety rules might be less exposed to the federal executive order, which suggested leaving that domain to the states. [27]

The Dormant Commerce Clause is the most likely federal-constitutional theory for any future challenge. SB 53 applies to any frontier developer doing business in California, which in practice means any company that wants to operate in the state's market, and so a challenger would need to argue that the law has the practical effect of regulating extraterritorial conduct. The Future of Privacy Forum has argued that SB 53's structure is more vulnerable on this point than New York's later RAISE Act, because the New York statute is explicitly tied to operations within New York, whereas SB 53 attaches to any frontier developer doing business in California. The First Amendment is a secondary theory, since the framework and transparency report requirements are compelled speech. xAI's parallel challenge to AB 2013, California's training-data transparency law, could provide a template for any future SB 53 challenge. [4][5][35]

## Comparable state laws

With no comprehensive federal AI law in effect, several U.S. states moved during 2025 and 2026 to fill the gap. SB 53 was the first of the states to adopt a frontier-model-specific safety regime, but it was followed in short order by New York's RAISE Act and amendments to other state laws. The principal comparables are summarised below.

| State | Law | Effective | Scope | Principal mechanism |
|---|---|---|---|---|
| California | SB 53 (TFAIA) | January 1, 2026 | Frontier developers (>10^26 FLOPs); strict duties at >$500M revenue | Framework + transparency reports + incident reporting + whistleblower; AG enforcement; civil penalty up to $1M per violation |
| New York | Responsible AI Safety and Education (RAISE) Act | January 1, 2027 (signed December 2025; March 27, 2026 chapter amendments aligned it with TFAIA) [52] | Frontier developers operating in New York | Transparency-and-incident regime; civil penalty up to $1M (first) and $3M (subsequent); 72-hour incident reporting; no whistleblower provisions |
| Colorado | Colorado AI Act (SB 24-205), replaced by SB 26-189 | Never took effect: enforcement was stayed in April 2026 and the narrower SB 26-189, effective January 1, 2027, replaced it in May 2026 [50][51] | High-risk AI systems used for consequential decisions | Algorithmic discrimination duty of reasonable care; impact assessments; consumer disclosures |
| Texas | Responsible AI Governance Act (TRAIGA, HB 149) | January 1, 2026 | Broad: bans certain harmful uses; AI sandbox programme | Prohibitions on specified harms; 36-month regulatory sandbox under Department of Information Resources |
| Utah | AI Policy Act and amendments (SB 149, SB 226, others) | 2024-2026, rolling | Consumer-facing AI uses | Disclosure of AI interaction; sectoral rules in healthcare, mental health, and education |

SB 53 is the only one of these laws to focus exclusively on frontier-scale models. Colorado's Colorado AI Act addresses a different policy problem (algorithmic discrimination in consequential decisions). Texas' TRAIGA adopts a much broader scope but uses prohibitions on specified harms rather than disclosure obligations. Utah's regime is overwhelmingly disclosure-based and consumer-facing. Of the comparators, only New York's RAISE Act is a direct sister statute, and the New York legislature explicitly amended its bill in March 2026 to track TFAIA's structure more closely. [5][20][35][36]

## What is the implementation status of SB 53?

The substantive transparency, framework, and incident reporting provisions of SB 53 took effect on January 1, 2026. As of mid-2026, the implementation picture was still developing.

Anthropic published its Frontier Compliance Framework on December 19, 2025 as its SB 53 compliance document, keeping its Responsible Scaling Policy (updated to version 3.0 on February 24, 2026) as a separate voluntary policy. Google DeepMind had published the third version of its Frontier Safety Framework on September 22, 2025. OpenAI's relevant documents include its Preparedness Framework (version 2, April 2025) and the system cards published with [GPT-5](https://aiwiki.ai/wiki/gpt-5), [GPT-5.1](https://aiwiki.ai/wiki/gpt-5.1), [GPT-5.2](https://aiwiki.ai/wiki/gpt-5.2), and [GPT-5.5](https://aiwiki.ai/wiki/gpt-5.5); the statute treats information published in a system card as satisfying the transparency report requirement. xAI published a redrafted Frontier Artificial Intelligence Framework on December 31, 2025, replacing the August 2025 Risk Management Framework. [1][24][28][29]

The California Office of Emergency Services set up a TFAIA reporting portal, run through its California Cybersecurity Integration Center, through which frontier developers and members of the public can confidentially submit critical safety incident reports and large frontier developers can submit periodic summaries of internal-use catastrophic risk assessments; it also published a worksheet of the portal's fields. [48] The CalCompute framework report is due January 1, 2027, but the entire CalCompute section becomes operative only upon a budget appropriation. The California Attorney General had not, as of May 2026, brought any enforcement action under SB 53. No critical safety incidents had been publicly reported as of May 2026. [3][4][48]

METR's December 2025 update of its survey of common elements in frontier AI safety policies, which by then covered twelve companies, added references to SB 53, and the Stanford CRFM Foundation Model Transparency Index published a December 2025 edition. The [Frontier Model Forum](https://aiwiki.ai/wiki/frontier_model_forum)'s technical report series on frontier safety frameworks served as the principal industry-led standardisation venue, publishing reports on risk taxonomy, mitigations, and third-party assessments during 2025 and 2026 that explicitly referenced SB 53. [38][39][40]

### First-quarter 2026 compliance behaviour by labs

The first published compliance documents came around the law's effective date. Anthropic's Frontier Compliance Framework describes how the company assesses and mitigates cyber offense and chemical, biological, radiological, and nuclear risks as well as the risks of AI sabotage and loss of control, sets out a tiered system for evaluating model capabilities, and covers model weight protection and incident response. xAI's Frontier Artificial Intelligence Framework, dated December 2025, states that it complies with the TFAIA. [24][29]

The statute settles the treatment of system cards: Section 22757.12(c)(3) deems a developer compliant if it publishes the required transparency report information as part of a larger document, including a system card or [model card](https://aiwiki.ai/wiki/model_card). The Department of Technology's first recommendations under Section 22757.14, due on or before January 1, 2027, concern the definitions of frontier model, frontier developer, and large frontier developer rather than the format of transparency reports. [1]

### Office of Emergency Services and Department of Technology readiness

By statute, the OES mechanism must collect the date of the incident, the reasons it qualifies as a critical safety incident, a short and plain statement describing it, and whether it was associated with internal use of a frontier model; large frontier developers file their internal-use risk summaries through the same portal every three months or on another reasonable schedule. [1][48]

## Has SB 53 been challenged in court?

As of May 2026, no court had enjoined or otherwise paused the operation of SB 53. xAI's earlier challenge to California's AB 2013 (the AI training data transparency law of 2024) is the closest existing analogue, raising First Amendment compelled speech and trade secret arguments that could be imported into a future SB 53 challenge. The Ninth Circuit's November 18, 2025 grant of an injunction against California's separate climate disclosure law SB 261 was widely cited as evidence that compelled disclosure regimes face meaningful federal-constitutional risk, although SB 261 implicates different statutory and First Amendment frameworks than SB 53. [35][41]

The Trump administration's December 11, 2025 executive order on state AI regulation directed the Department of Justice to consider preemption and Dormant Commerce Clause challenges to state AI laws, but through September 2026 no such suit had been filed against SB 53; the Task Force's first reported action was in the Colorado case. [50] Carnegie Endowment analysts writing in December 2025 said most legal commentators expected the administration to face an uphill battle in court, and predicted that the order's most likely outcome would be a federal incident reporting system. [33][34]

## See also

- [California SB 1047](https://aiwiki.ai/wiki/sb_1047)
- [AI regulation](https://aiwiki.ai/wiki/ai_regulation)
- [EU AI Act](https://aiwiki.ai/wiki/eu_ai_act)
- [Executive Order on AI](https://aiwiki.ai/wiki/ai_executive_order)
- [AI Action Plan](https://aiwiki.ai/wiki/ai_action_plan)
- [Responsible Scaling Policy](https://aiwiki.ai/wiki/responsible_scaling_policy)
- [AI safety](https://aiwiki.ai/wiki/ai_safety)
- [AI Safety Summit](https://aiwiki.ai/wiki/ai_safety_summit)
- [Bletchley Declaration](https://aiwiki.ai/wiki/bletchley_declaration)
- [Seoul Declaration](https://aiwiki.ai/wiki/seoul_declaration)
- [Frontier Model Forum](https://aiwiki.ai/wiki/frontier_model_forum)
- [AI Safety Institute](https://aiwiki.ai/wiki/ai_safety_institute)
- [Scott Wiener](https://aiwiki.ai/wiki/scott_wiener)
- [Gavin Newsom](https://aiwiki.ai/wiki/gavin_newsom)

## References

1. California Legislative Information, "SB-53 Artificial intelligence models: large developers," enrolled text, September 17, 2025. https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB53
2. California Legislative Information, "SB-53 Bill History," 2025-2026 Regular Session. https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202520260SB53
3. Office of Governor Gavin Newsom, "Governor Newsom signs SB 53, advancing California's world-leading artificial intelligence industry," September 29, 2025. https://www.gov.ca.gov/2025/09/29/governor-newsom-signs-sb-53-advancing-californias-world-leading-artificial-intelligence-industry/
4. WilmerHale Privacy and Cybersecurity Law, "Transparency in Frontier Artificial Intelligence Act (SB-53): California Requires New Standardized AI Safety Disclosures," October 1, 2025. https://www.wilmerhale.com/en/insights/blogs/wilmerhale-privacy-and-cybersecurity-law/20251001-transparency-in-frontier-artificial-intelligence-act-sb-53-california-requires-new-standardized-ai-safety-disclosures
5. Future of Privacy Forum, "California's SB 53: The First Frontier AI Law, Explained." https://fpf.org/blog/californias-sb-53-the-first-frontier-ai-law-explained/
6. Joint California Policy Working Group on AI Frontier Models, "The California Report on Frontier AI Policy," June 17, 2025. https://www.gov.ca.gov/wp-content/uploads/2025/06/June-17-2025-%E2%80%93-The-California-Report-on-Frontier-AI-Policy.pdf
7. Brookings Institution, "What is California's AI safety law?" 2025. https://www.brookings.edu/articles/what-is-californias-ai-safety-law/
8. Carnegie Endowment for International Peace, "California Just Passed the First U.S. Frontier AI Law. Here's What It Does," October 2025. https://carnegieendowment.org/emissary/2025/10/california-sb-53-frontier-ai-law-what-it-does
9. Anthropic, "Anthropic is endorsing SB 53," September 8, 2025. https://www.anthropic.com/news/anthropic-is-endorsing-sb-53
10. SF Standard, "Newsom signs AI safety law opposed by Meta, Google, OpenAI," September 29, 2025. https://sfstandard.com/2025/09/29/gavin-newsom-california-ai-legislation-law-technology/
11. NPR, "California Gov. Newsom vetoes AI safety bill that divided Silicon Valley," September 29, 2024. https://www.npr.org/2024/09/29/nx-s1-5119792/newsom-ai-bill-california-sb1047-tech
12. Wikipedia, "Safe and Secure Innovation for Frontier Artificial Intelligence Models Act." https://en.wikipedia.org/wiki/Safe_and_Secure_Innovation_for_Frontier_Artificial_Intelligence_Models_Act
13. Joint California Policy Working Group on AI Frontier Models, "The California Report on Frontier AI Policy," arXiv preprint. https://arxiv.org/abs/2506.17303
14. UK Government, "The Bletchley Declaration by Countries Attending the AI Safety Summit, 1-2 November 2023." https://www.gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration/the-bletchley-declaration-by-countries-attending-the-ai-safety-summit-1-2-november-2023
15. UK Government, "Frontier AI Safety Commitments, AI Seoul Summit 2024." https://www.gov.uk/government/publications/frontier-ai-safety-commitments-ai-seoul-summit-2024/frontier-ai-safety-commitments-ai-seoul-summit-2024
16. White House, "Winning the Race: America's AI Action Plan," July 23, 2025. https://www.whitehouse.gov/wp-content/uploads/2025/07/Americas-AI-Action-Plan.pdf
17. Latham and Watkins, "California Assumes Role as Lead US Regulator of AI," October 2025. https://www.lw.com/en/insights/california-assumes-role-as-lead-us-regulator-of-ai
18. Mayer Brown, "California Enacts SB-53, Creating New Requirements for Developers of Frontier Artificial Intelligence Models and Related Whistleblower Provisions," October 2025. https://www.mayerbrown.com/en/insights/publications/2025/10/california-enacts-sb-53-creating-new-requirements-for-developers-of-frontier-artificial-intelligence-models-and-related-whistleblower-provisions
19. Greenberg Traurig LLP, "California Expands Whistleblower Retaliation Protections for Employees in the AI Sector," October 2025. https://www.gtlaw.com/en/insights/2025/10/california-expands-whistleblower-retaliation-protections-for-employees-in-the-ai-sector
20. Fisher Phillips LLP, "California Lawmakers Pass Landmark AI Transparency Law for Frontier Models: How SB 53 Differs from Last Year's Failed Attempt," 2025. https://www.fisherphillips.com/en/insights/insights/california-lawmakers-pass-landmark-ai-transparency-law-for-frontier-models
21. LegiScan, "California SB53, 2025-2026 Regular Session," history and votes. https://legiscan.com/CA/bill/SB53/2025
22. CalMatters Digital Democracy, "SB 53: Artificial intelligence models: large developers." https://calmatters.digitaldemocracy.org/bills/ca_202520260sb53
23. California State Senate, "Governor Newsom Signs Senator Wiener's Landmark AI Law To Set Commonsense Guardrails, Boost Innovation," September 29, 2025. https://sd11.senate.ca.gov/news/governor-newsom-signs-senator-wieners-landmark-ai-law-set-commonsense-guardrails-boost
24. Anthropic, "Sharing our compliance framework for California's Transparency in Frontier AI Act," December 19, 2025. https://www.anthropic.com/news/compliance-framework-SB53
25. NBC News, "Anthropic backs California's SB 53 AI bill," September 8, 2025. https://www.nbcnews.com/tech/tech-news/anthropic-backs-californias-sb-53-ai-bill-rcna229908
26. TechCrunch, "California Governor Newsom signs landmark AI safety bill SB 53," September 29, 2025. https://techcrunch.com/2025/09/29/california-governor-newsom-signs-landmark-ai-safety-bill-sb-53/
27. Carnegie Endowment for International Peace, "AI's New Test on California's Direct Democracy," December 2025. https://carnegieendowment.org/emissary/2025/12/openai-california-ai-ballot-direct-democracy
28. Google DeepMind, "Strengthening the Frontier Safety Framework," September 22, 2025. https://deepmind.google/blog/strengthening-our-frontier-safety-framework/
29. xAI, "xAI Frontier Artificial Intelligence Framework" (PDF), December 31, 2025. https://data.x.ai/2025-12-31-xai-frontier-artificial-intelligence-framework.pdf
30. xAI, "xAI Risk Management Framework" (PDF), August 20, 2025. https://data.x.ai/2025-08-20-xai-risk-management-framework.pdf
31. Encode AI, statements on SB 53 signing, September 29, 2025. https://encodeai.org/
32. Wharton AI and Analytics Initiative, "SB 53: What California's New AI Safety Law Means for Developers." https://ai-analytics.wharton.upenn.edu/wharton-accountable-ai-lab/sb-53-what-californias-new-ai-safety-law-means-for-developers/
33. Paul Hastings LLP, "President Trump Signs Executive Order Challenging State AI Laws," December 2025. https://www.paulhastings.com/insights/client-alerts/president-trump-signs-executive-order-challenging-state-ai-laws
34. Carnegie Endowment for International Peace, "The Most Likely Outcomes of Trump's Order Targeting State AI Laws," December 2025. https://carnegieendowment.org/emissary/2025/12/ai-trump-order-state-law-what-to-know
35. Future of Privacy Forum, "The RAISE Act vs. SB 53: A Tale of Two Frontier AI Laws," 2026. https://fpf.org/blog/the-raise-act-vs-sb-53-a-tale-of-two-frontier-ai-laws/
36. King and Spalding, "New State AI Laws are Effective on January 1, 2026, But a New Executive Order Signals Disruption." https://www.kslaw.com/news-and-insights/new-state-ai-laws-are-effective-on-january-1-2026-but-a-new-executive-order-signals-disruption
37. Nelson Mullins, "California SB 53 Expanded Compliance Guide for Frontier AI Developers." https://www.nelsonmullins.com/insights/blogs/ai-task-force/ai/california-sb-53-expanded-compliance-guide-for-frontier-ai-developers
38. METR, "Common Elements of Frontier AI Safety Policies (December 2025 Update)." https://metr.org/blog/2025-12-09-common-elements-of-frontier-ai-safety-policies/
39. Stanford CRFM, Foundation Model Transparency Index, December 2025. https://crfm.stanford.edu/fmti/December-2025/
40. Frontier Model Forum, technical report series on frontier AI safety frameworks. https://www.frontiermodelforum.org/technical-reports/
41. Harvard Law School Forum on Corporate Governance, "California Climate Disclosure Law SB 261 Implementation Halted: Ninth Circuit Grants Injunction Pending Appeal," December 3, 2025. https://corpgov.law.harvard.edu/2025/12/03/california-climate-disclosure-law-sb-261-implementation-halted-ninth-circuit-grants-injunction-pending-appeal/
42. WinBuzzer, "How California's SB 53 Law Affects AI Lab Whistleblowers (OpenAI policy update)," January 28, 2026. https://winbuzzer.com/2026/01/28/openai-updates-whistleblower-policy-california-law-xcxwbn/
43. Introl, "Federal vs State AI Law Showdown: Trump Executive Order 2026." https://introl.com/blog/federal-state-ai-law-showdown-trump-executive-order-2026
44. EDUCAUSE Review, "Trump Signs AI Executive Order as States Implement AI Laws," March 2026. https://er.educause.edu/articles/2026/3/trump-signs-ai-executive-order-as-states-implement-ai-laws
45. SF Examiner, "California AI safety bill becoming template for other states." https://www.sfexaminer.com/news/technology/california-ai-safety-bill-becoming-template-for-other-states/article_34a97de6-0ac7-4b75-9fe4-1ade33b1bbe7.html
46. Goodwin Procter, "California Moves to Regulate Frontier AI With a Focus on Catastrophic Risk," November 2025. https://www.goodwinlaw.com/en/insights/publications/2025/11/alerts-technology-aiml-california-moves-to-regulate-frontier-ai-with-a-focus-on-catastrophic-risk
47. Fortune, "California's landmark AI safety law SB 53," September 30, 2025. https://fortune.com/2025/09/30/california-landmark-ai-safety-law-sb-53-newsom-signs-silicon-valley-protect-whistleblowers/
48. California Governor's Office of Emergency Services, "Transparency in Frontier AI Act Reporting." https://www.caloes.ca.gov/office-of-the-director/operations/homeland-security/california-cybersecurity-integration-center/transparency-in-frontier-ai-act-reporting/
49. The Next Web, "OpenAI wants California to toughen the AI law it once fought," August 2026. https://thenextweb.com/news/openai-sb-53-california-ai-safety-law-amendments
50. Fortune, "After a decade of failed bills and three years of resignations, Washington finally discovers it cares about AI safety," September 12, 2026. https://fortune.com/2026/09/12/after-three-years-of-silence-washington-discovers-it-finally-cares-about-ai-safety/
51. CBS News Colorado, "Colorado governor signs new AI regulation bill to replace 2024 law," May 14, 2026. https://www.cbsnews.com/colorado/news/ai-regulation-colorado-governor-signs-law/
52. Cooley, "New York's Frontier AI Law Gets a California Makeover, With Some Key Differences," March 31, 2026. https://www.cooley.com/news/insight/2026/2026-03-31-new-yorks-frontier-ai-law-gets-a-california-makeover-with-some-key-differences