Separating AI hype from reality.

California Proposes 48-Hour AI Incident Reporting and State Shutdown Authority, But Nothing Is Law Yet

November 19, 2026 · Regulation · California

Governor Gavin Newsom on November 18 released a set of legislative recommendations that would give California officials power to order emergency shutdowns of frontier AI models, cut critical safety incident reporting deadlines from 15 days to 48 hours, and impose continuous independent oversight on large AI developers. The proposals are not enacted law; they are intended to inform bills when the Legislature returns December 7 [1][2][4].

The 85-page report, AI Safety in 2026 and Beyond, was produced by the California Government Operations Agency with the Office of Emergency Services and four outside advisers: Jason Goldman, Gillian Hadfield, Alondra Nelson, and Rob Reich [1][2][3][4]. It responds to a September 18 executive order in which Newsom asked for proposals on independent oversight and an “AI kill switch” within two months [7].

What the recommendations would do

Emergency shutdown authority. State officials could order an immediate shutdown when they have “credible reason to believe there is an imminent catastrophic risk or major loss-of-control event and the developer refuses to contain the danger.” The order must be the “narrowest workable intervention,” takes effect immediately, and stays in force during judicial review, though developers could seek a temporary restraining order [2][4]. The power would apply to companies such as OpenAI and Anthropic [1][4].

Faster incident reporting. The deadline for most critical safety incident reports would drop from 15 days to 48 hours, while the existing 24-hour requirement for imminent catastrophic risks would remain. Reporting would expand to cover attempted model-weight theft, all internal or external loss-of-control incidents, and a new near-miss channel [2][4]. Whistleblower protections would extend to contractors and former employees [1][4].

Continuous independent oversight. Every large frontier developer would be monitored by state-licensed Independent Verification Organizations (IVOs). The Government Operations Agency would license IVOs and could revoke licenses for insufficient expertise, independence, or rigor. The state—not the AI companies—would assign evaluations to auditors. Auditors would get access to models, personnel, systems, training environments, and records, and send unredacted assessments to the state [2][4].

Risk Containment Protocols. Each large frontier developer would maintain a protocol covering compute, agents, internal access, and deployment, verified by an IVO and certified by a senior company official. Testing would include failure cases such as a model copying its own weights to resist containment. Open-weight developers would be responsible only for deployments they control [2][4].

Recursive self-improvement (RSI) caps. Officials could impose a time-limited, periodically reviewed cap on automated AI research when it exceeds a developer’s committed thresholds or advances faster than its safeguards and evaluations, creating a material risk of catastrophic harm. The cap would be lifted after an IVO verified adequate safeguards were in place [2].

Automated AI research reporting. Companies automating AI R&D would report quarterly how their compute is divided among training, research experiments, inference for AI-research agents, safety and security work, external customers, and unattributed activity, with IVOs verifying the categories [2].

Capability-based frontier definition. Models capable of actions that could create catastrophic risk would count as frontier models for reporting purposes even if trained with fewer than 10^26 integer or floating-point operations, addressing the concern that relying on training scale alone has left dangerous incidents outside state reporting requirements [2].

CalCompute funding. A surcharge on the verification fees developers pay to auditors would help finance CalCompute, the state’s public cloud initiative, with some capacity reserved for independent safety research and public oversight [2].

Political context and next steps

Newsom framed the effort as filling a vacuum left by federal inaction. “Instead of protecting the public, Donald Trump shrugs his shoulders and leaves AI regulation up to the companies themselves,” he said in a press release [1]. Governor-elect Xavier Becerra said in a written statement that he will move quickly with the Legislature to put “full incident reporting and real shutdown authority into law” [1][4].

Lawmakers return to Sacramento on December 7, when bills based on the recommendations can be introduced [1][4]. State Senator Scott Wiener, who wrote a prior state law on catastrophic AI dangers and was recently elected to Congress, has pledged to introduce AI safety regulation after he is sworn in [1].

The recommendations build on existing California laws: SB 53 (2025), the Transparency in Frontier Artificial Intelligence Act, which requires frontier AI developers to disclose safety frameworks, report critical safety incidents, and protect whistleblowers; SB 813 and AB 1405 (2026), which establish a framework for certifying independent verification organizations and a state registry for AI auditors [6][7].

Industry reaction

Anthropic’s chief global affairs officer, Tino Cuéllar, said the company is “encouraged to see the report move the conversation forward on issues like the independent evaluation of AI models” and will “continue working with leaders in California and throughout the world to make sure we keep making progress on these issues at a critical time” [1][4]. No reaction from OpenAI was reported [4].

What the report does not cover

The report focuses only on catastrophic risk—defined in state law as a single event capable of killing more than 50 people or causing $1 billion in theft or property damage through a cyber attack or autonomous system activity [1]. It does not explore protections from other AI harms such as automated discrimination, job loss, or threats to children’s safety or mental health [1]. Coauthors argue that building a system for credentialed auditors to test models for catastrophic risk will make it easier to evaluate models for other forms of harm [1].

From proposal to law

The recommendations are expert proposals, not binding requirements. The September 18 executive order accelerated implementation timelines for SB 813 and AB 1405 and convened the expert panel [7]. The November 18 report is the panel’s output. Whether the Legislature enacts any, all, or none of the recommendations—and on what timeline—remains an open question.

Sources

  1. Newsom touts AI kill switch to halt dangerous model releases (calmatters.org)
  2. Newsom Proposes Mandatory Frontier AI Oversight and Shutdown ... (unite.ai)
  3. California Gets Power to Shut Down Dangerous AI Models (emergent.sh)
  4. California report proposes AI emergency shutdown power and 48- ... (aichatdaily.com)
  5. Newsom Touts an AI Kill Switch to Halt Dangerous OpenAI, Anthropic ... (ground.news)
  6. Governor Newsom’s nation-leading AI safety framework is getting even stronger as California releases next steps on mandatory safety standards and oversight | Governor of California (gov.ca.gov)
  7. Governor Newsom issues executive order to accelerate independent oversight and advance the creation of an AI kill switch | Governor of California (gov.ca.gov)