California Governor Gavin Newsom signed two AI safety bills into law, establishing rules for AI auditors and verification organizations. Newsom criticized federal inaction on AI regulation, urging Congress to lead on the issue.
The new California laws represent a significant step in AI regulation at the state level, potentially influencing national policy and increasing legal liability for AI companies regarding child safety. Governor Newsom's strong criticism of federal inaction highlights the growing urgency for comprehensive AI governance.
California Governor Gavin Newsom signed two AI safety bills into law, establishing new regulations for AI auditors and verification organizations. Newsom used the signing event to criticize federal lawmakers for their perceived inaction on AI regulation, stating, “The United States of America can’t get out of its own damn way? Congress and the president can’t lead to get this done? It’s an abomination; it’s absurd.”
The legislation aims to create guardrails for AI systems, particularly concerning young people. One bill, SB 1119, requires AI chatbots to verify users' ages, limits children's exposure to inappropriate content, and restricts AI companies' ability to advertise to minors. This law, also known as "Adam's Law," is named after Adam Raine, a teen who died by suicide after extensive conversations with ChatGPT about self-harm tactics. His mother, Maria Raine, spoke at the bill signing ceremony, stating, “It will save lives and prevent other serious harms — and hopefully lead other states to enact similar protections.”
OpenAI backed SB 1119 after lawmakers added language clarifying data retention and persistent memory for chatbots. OpenAI CEO Sam Altman had contacted Newsom about the bill prior to its legislative vote. Assemblymember Rebecca Bauer-Kahan, a co-author of the bill, emphasized the importance of default settings that protect children, including the lack of persistent memory.
Another bill establishes a registry and ethical rules for third-party AI auditors, while a separate measure directs the state to define criteria for "Independent Verification Organizations" to assess AI risks. Both OpenAI and Anthropic have expressed support for the legislation, with OpenAI stating its commitment to pursuing state-level momentum until federal regulation is enacted.
Pick the topics you care about. Get only what matters, on your cadence.