California Lawmaker Revives Push for Mandatory AI Safety Reports

California's Renewed Push for AI Transparency

California State Senator Scott Wiener has reignited the debate on AI regulation with new amendments to Senate Bill 53 (SB 53). The bill aims to mandate the publication of safety and security protocols by major AI companies, as well as require reporting of safety incidents 1

1

. If passed, this legislation would make California the first state to impose significant transparency requirements on leading AI developers, potentially affecting industry giants like OpenAI, Google, Anthropic, and xAI 1

1

.

Background and Previous Attempts

This move follows the veto of Senator Wiener's previous bill, SB 1047, which faced strong opposition from Silicon Valley. Governor Gavin Newsom vetoed the bill and subsequently established a policy group of AI leaders to set goals for the state's AI safety efforts 1

1

. The amendments to SB 53 have been heavily influenced by the recommendations of this policy group, which emphasized the need for "requirements on industry to publish information about their systems" 1

1

.

Key Provisions of SB 53

The amended bill includes several notable provisions:

1. Mandatory Safety Reports: Companies developing AI models above a certain computing performance threshold would be required to publicly release safety and security protocols 2

   2

   .

2. Incident Reporting: AI companies would need to report any "critical safety incidents" to the state attorney general 2

   2

   .

3. Whistleblower Protections: The bill offers protection for employees of AI labs who believe their company's technology poses a "critical risk" to society 1

   1

   .

4. Public Computing Resource: SB 53 proposes the creation of CalCompute, a public cloud computing cluster to support startups and researchers in developing large-scale AI 1

   1

   .

Industry Response and Implications

The AI industry has shown mixed reactions to transparency requirements. While companies like Anthropic have broadly endorsed the need for increased transparency, others such as OpenAI, Google, and Meta have been more resistant 1

1

. The bill represents a toned-down version of previous AI safety proposals but could still force AI companies to disclose more information than they currently do 1

1

.

National Context and Similar Efforts

California's push for AI regulation is not isolated. In New York, Governor Kathy Hochul is considering a similar AI safety bill, the RAISE Act 1

1

. These state-level efforts gained momentum after a federal proposal for a 10-year AI moratorium on state AI regulation failed in a 99-1 Senate vote 1

1

.

The Road Ahead for SB 53

The amended SB 53 now heads to the California State Assembly Committee on Privacy and Consumer Protection for approval. If successful, it will need to pass through several other legislative bodies before reaching Governor Newsom's desk 1

1

. Senator Wiener has expressed openness to further refinements, stating, "The bill continues to be a work in progress, and I look forward to working with all stakeholders in the coming weeks to refine this proposal into the most scientific and fair law it can be" 1

1

.

As the debate over AI regulation continues, the fate of SB 53 will be closely watched by both the tech industry and policymakers across the nation. Its success or failure could set a precedent for how states approach the challenge of ensuring AI safety and transparency in the absence of comprehensive federal regulation.