**California Law Sets Precedent with Mandatory AI Safety Disclosures**
*By Mudit Dube | Sep 30, 2025, 01:59 PM*
—
**What’s the Story?**
California Governor Gavin Newsom has signed a landmark bill, SB 53, into law. This groundbreaking legislation requires major artificial intelligence (AI) companies, including giants like OpenAI and Google, to disclose how they plan to mitigate potential significant risks posed by their advanced AI models. The move underscores California’s commitment to lead regulation efforts in an industry crucial to its economy.
—
**Regulatory Approach**
In a press release, Governor Newsom emphasized California’s ability to craft regulations that both protect communities and foster the growth of the AI industry. The new law addresses a critical gap left by the U.S. Congress, which has yet to enact comprehensive AI legislation.
Newsom further stated that should federal AI standards eventually be developed, California lawmakers would work to align state regulations with those standards while maintaining the high bar set by SB 53.
—
**Compliance Requirements**
SB 53 targets companies with annual revenues exceeding $500 million. These companies must conduct risk assessments focused on the potential misuse of their technology—such as unintended behaviors or use in developing bioweapons. These assessments are to be made publicly available.
The law also introduces fines of up to $1 million per violation, ensuring strict compliance.
Jack Clark, co-founder of the AI company Anthropic, praised SB 53 as “a strong framework that balances public safety with continued innovation.”
—
**Industry Pushback**
Despite positive responses from some experts, the law has faced criticism from segments of the tech industry. Several tech firms have expressed concerns that state-level AI regulations could result in a “patchwork of regulation” that might hinder innovation and complicate compliance.
Collin McCune, head of government affairs at Andreessen Horowitz, warned that SB 53 “could set a precedent for states to lead in governing the national AI market,” which some see as potentially problematic.
—
**Potential Preemption and Federal Discussions**
At the federal level, California Republican US Representative Jay Obernolte is drafting AI legislation that may preempt certain state laws, including SB 53. Meanwhile, Democrats in Congress are also actively discussing frameworks for establishing nationwide AI standards.
During a recent House hearing on AI legislation, Representative Ted Lieu posed the question of whether regulation should remain at the state level or be centralized through Congress—highlighting the ongoing debate over the best path forward for AI governance in the U.S.
—
**Conclusion**
California’s SB 53 marks a significant step in AI regulation, setting a precedent for mandatory safety disclosures from leading AI companies. As debates continue over how to balance innovation with public safety, the law may influence both state and federal policymaking in the rapidly evolving AI landscape.
https://www.newsbytesapp.com/news/science/california-mandates-ai-safety-disclosures-with-sb-53-law/story
