What happened
Illinois' House of Representatives passed Senate Bill 315, requiring frontier AI labs, including OpenAI, Anthropic, and Google DeepMind, to undergo independent third-party safety audits. Governor JB Pritzker plans to sign the bill, which AI safety experts state would establish the nation's strongest check on major AI companies. This legislation mandates external verification of safety practices, moving beyond self-reporting requirements in other states like California and New York. OpenAI publicly supported the measure.
Why it matters
Independent audits introduce external accountability for frontier AI safety claims, shifting the burden of proof from self-reporting to verified compliance. This means AI labs must now demonstrate adherence to their safety standards through external validation. The mechanism is third-party verification, a new constraint on AI labs' self-attestation. This follows earlier state-level AI regulation efforts in Michigan and Alberta, indicating a growing trend of sub-national policy development in the absence of federal action.
