URL: gov.ca.gov Date: 2025-09-29 Class: foundational (primary-text from the California Governor's Office)
Summary
This is the official signing statement issued by the California Governor's Office on September 29, 2025, when Governor Gavin Newsom signed SB 53, the Transparency in Frontier Artificial Intelligence Act (TFAIA), authored by Senator Scott Wiener (D-San Francisco). The statement supplements the bill text with the executive branch's framing of the law (described as "commonsense guardrails"), the working-group lineage that produced the policy report informing the bill, and a five-part structure (transparency, innovation, safety, accountability, and responsiveness) by which the law will be administered. It also recites a set of statistics on California's position in the AI industry.
Statutory structure described in the statement
The signing statement organizes SB 53 around five elements:
- Transparency. Large frontier developers must publicly publish a framework on their website describing how they have incorporated national standards, international standards, and industry-consensus best practices.
- Innovation. The statement establishes the CalCompute consortium within the Government Operations Agency to "develop a framework for creating a public computing cluster." Per the statement, CalCompute is intended to "advance the development and deployment of artificial intelligence that is safe, ethical, equitable, and sustainable by fostering research and innovation."
- Safety. The law creates a mechanism for frontier AI companies and the public to report potential critical safety incidents to California's Office of Emergency Services.
- Accountability. Whistleblower protections cover those disclosing significant health and safety risks, with a civil penalty enforced by the Attorney General.
- Responsiveness. The California Department of Technology is directed to annually recommend appropriate updates based on multistakeholder input, technological developments, and international standards.
The bill text and the signing statement agree that CalCompute is housed under the Government Operations Agency. As of May 2026, the bill provided no funding level or chip-acquisition target for CalCompute.
Working-group lineage
The signing statement traces the bill to a deliberative policy process rather than presenting it as legislative will alone. It names the Joint California AI Policy Working Group, convened in fall 2024 at Newsom's request following his veto of SB 1047, and its March 2025 first-in-the-nation report, described as an empirical, science-based analysis of frontier-model capabilities and risks. The statement attributes the "trust but verify" framing to working-group member Cuéllar.
The working-group members named in the statement are:
- Mariano-Florentino (Tino) Cuéllar — former California Supreme Court Justice and former member of the National Academy of Sciences Committee on the Social and Ethical Implications of Computing Research. (See Mariano-Florentino Cuéllar.)
- Dr. Fei-Fei Li — Co-Director, Stanford Institute for Human-Centered Artificial Intelligence. (See Li Fei-Fei.)
- Jennifer Tour Chayes — Dean, College of Computing, Data Science, and Society at UC Berkeley. (See Jennifer Tour Chayes.)
The statement positions SB 53 as the output of this working group's report. Whether the working group informed the bill text is supported by the direct lineage the statement describes (the Cuéllar/Li/Chayes report preceding the bill), and the "trust but verify" principle is quoted directly from Cuéllar.
California AI-industry statistics cited
The statement recites several figures on California's standing in the AI sector:
- 32 of the top 50 AI companies worldwide (Forbes AI50) are California-based.
- California accounted for 15.7% of all U.S. AI job postings in 2024, ranking first by state, ahead of Texas (8.8%) and New York (5.8%) (Stanford AI Index 2025).
- More than half of global VC AI/ML funding went to Bay Area firms in 2024 (PitchBook).
- Three of the four companies valued above $3 trillion (Google, Apple, and Nvidia) are California-based and AI-involved.
These figures track standard 2025 industry rankings.
Framing and reception
Newsom's statement characterizes SB 53 as striking a balance between innovation and "commonsense guardrails." Wiener, the bill's author, described the law as one that "strikes that balance" and stated that "SB 53 fills [the gap of federal inaction] and presents a model for the nation to follow." The "model for the nation" characterization is Wiener's framing; the NY RAISE Act, signed December 22, 2025, followed the SB 53 model, and its own signing statement (Governor Hochul Signs Nation-Leading Legislation to Require AI Frameworks for AI Frontier Models (NY DFS, December 22 2025)) says New York "builds on California's recently adopted framework." That sequence positions SB 53 (signed September 29, 2025) and the RAISE Act as a paired California–New York state frontier-transparency effort.
The statement's "balance" framing stands against Newsom's earlier veto of SB 1047. Both SB 1047 and SB 53 were characterized in terms of balance, but with different operative provisions: SB 1047 included pre-training shutdown obligations and a penalty set at 30% of compute cost, whereas SB 53 relies on post-deployment disclosure and a penalty of $1 million per violation. Newsom vetoed SB 1047 partly citing compute-threshold concerns.
The statement asserts California's leadership in AI policy, a position that runs against the federal preemption push from OpenAI and Google and the Trump December 2025 state-preemption executive order (EO — Trump Federal Preemption of State AI Laws (Dec 11, 2025)), which would override state-level AI regulation.
Relationships
- deploys-in: California SB 53 (this is the signing statement primary text for that legislation page)
- supports: AI Pre-Release Vetting, AI Transparency, State-Level AI Regulation, California Effect
- contradicts: EO — Trump Federal Preemption of State AI Laws (Dec 11, 2025) (federal preemption push)
- depends-on: The California Report on Frontier AI Policy (the working-group report informing the bill), Anthropic's Responsible Scaling Policy (Version 3.1) (de facto standard for what "framework" means)
- related: Scott Wiener (bill author), Gavin Newsom (signer), Mariano-Florentino Cuéllar, Li Fei-Fei, Jennifer Tour Chayes (working-group), Governor Hochul Signs Nation-Leading Legislation to Require AI Frameworks for AI Frontier Models (NY DFS, December 22 2025) (sister state primary text), New York RAISE Act
- regulated-by: California Attorney General (enforcement); California Office of Emergency Services (incident reporting); California Department of Technology (annual updates)