Illinois Law Mandates AI Safety Testing for Anthropic and OpenAI
TL;DR. Illinois passed a state law requiring major AI firms like Anthropic and OpenAI to submit public safety plans and report critical incidents. - Governor J.B. Pritzker intends to sign the bill, which mandates independent, third-party safety testing of frontier models. - The law protects whistleblowers reporting emerging safety risks within AI companies. - Anthropic and OpenAI supported the legislation, viewing it as establishing a baseline for safety protocols.
- Illinois passed SB 315, a landmark AI safety law requiring large AI firms to submit public safety plans.
- The law mandates annual reports on independent, third-party safety testing of frontier AI models.
- Companies must report critical safety incidents within 72 hours, or 24 hours if there's imminent risk of harm.
- The legislation includes whistleblower protections for employees reporting AI safety risks.
- OpenAI and Anthropic supported the bill, implying they can easily meet its requirements.
- The law establishes a baseline for AI safety, benefiting larger firms over smaller ones seeking compliance.