Illinois just set the bar for state-level AI regulation, and it’s not a low one. Senate Bill 315, the Artificial Intelligence Safety Measures Act, sailed through the state Senate with a 52-5 vote, creating what amounts to the most aggressive AI oversight framework any US state has attempted.
Governor JB Pritzker has said he’ll sign it. When he does, companies building the most powerful AI systems in the world, think OpenAI, Anthropic, Google, Meta, and xAI, will face mandatory annual independent audits of their safety practices and be required to report on their models’ catastrophic risk capabilities.
What the bill actually does
The bill requires large AI developers to submit to third-party verification of their safety standards every year. Not a self-assessment. Not a blog post about “responsible AI.” An actual independent audit, conducted by someone who doesn’t have stock options in the company being reviewed.
It also mandates that these companies disclose the catastrophic risk capabilities of their models. If your AI system could theoretically be used to cause large-scale harm, you have to tell people about it, formally and on the record.
