Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Sb53 Reporting topic
No spam. Unsubscribe anytime.
Lawmakers press firms on SB 53 thresholds after AI incidents
Summary
Assembly members raised concerns that SB 53's reporting thresholds—loss of control tied to death/serious injury or $1 billion in damage—may leave the state without early warning of AI-driven incidents; OES and companies said they are evaluating thresholds and engaging with the state.
During the hearing, legislators pressed companies and state officials on whether the Transparency and Frontier Artificial Intelligence Act (SB 53) gives California adequate visibility into serious AI-related cybersecurity incidents.
One lawmaker said the bill's definition of "loss of control" triggers reporting only when an incident is "reasonably foreseeable to lead to the death or serious injury of 50 people or $1,000,000,000 in damage," and asked whether that leaves the state with gaps for less-catastrophic but still consequential events. Jonathan Snow of the Governor's Office of Emergency Services said OES is "working collaboratively with those AI companies" and evaluating the thresholds; he added the state will work with legislators to evolve the thresholds if needed.
Panelists urged more robust pre-release evaluation and broader disclosure practices than the current statutory minimums. Kyla Grew (Anthropic) and John Lindsay (OpenAI) described voluntary transparency measures and independent reviews, and recommended that legislators consider lowering thresholds or creating additional reporting pathways so that the state receives timely information even for incidents that fall below SB 53's statutory trigger.
