Taking a decisive step in the debate over how AI systems should be regulated, Illinois lawmakers have approved legislation that would impose some of the most extensive AI oversight requirements in the US.
Senate Bill 315 cleared the Illinois House by a 110-0 vote after previously passing the Senate 52-5. The measure now heads to Governor JB Pritzker, who has indicated he intends to sign it.
The legislation focuses on the largest developers of AI models and introduces requirements that extend beyond existing laws in California and New York. Most significant: it would require independent third-party reviews of AI safety practices, creating an external verification process that does not currently exist in comparable state laws.
The bill applies to major AI developers meeting specified thresholds, including companies with at least $500 million in annual revenue and significant computing resources used to train advanced models.
Under the legislation, covered companies would be required to publish detailed safety frameworks explaining how they assess model capabilities, evaluate severe risks, monitor incidents, and implement safeguards. Those frameworks would also need regular updates as systems evolve.
The bill requires outside auditors to evaluate whether companies are following their own published safety commitments. Supporters argue that independent verification is necessary because AI developers currently evaluate their own compliance without external review.
“Frontier developers now have to produce verifiable evidence of how they measure model capability, assess catastrophic risk, and respond to safety incidents, and the teams building products on those models inherit that documentation burden downstream,” said Mitch Ashley, VP and Practice Lead for AI-Native Software Engineering at The Futurum Group.
OpenAI and Anthropic Back the Legislation
Elected officials, tech leaders and citizens across the US have joined the debate about the absence of a federal regulatory framework for advanced AI systems. State governments in particular have become a venue for AI policy initiatives as Congress remains divided over how to address the technology’s opportunities and risks.
Illinois lawmakers said the bill is designed to address concerns surrounding highly capable AI models while avoiding restrictions that could slow innovation. Proponents have described the legislation as an effort to establish accountability measures before AI systems become even more powerful and widely deployed.
The proposal has attracted support from some of the industry’s leading AI companies. OpenAI and Anthropic publicly backed the legislation during its passage through the General Assembly.
Supporters argue that the legislation formalizes safety practices many leading AI labs already claim to follow voluntarily, including transparency reporting and incident disclosure processes.
The audit provision generated the most debate during the legislative process. Critics contend that the requirement could force companies to share sensitive information with outside organizations despite the absence of nationally recognized AI auditing standards.
Tech industry groups opposing the measure argued that independent reviewers may be required to make subjective judgments without clear federal guidelines governing AI safety evaluations. Lawmakers responded to those concerns through a series of amendments. Revisions clarified auditor qualifications, established requirements for protecting proprietary data, and refined expectations for audit procedures.
Additional changes removed any private right of action, preventing individuals from filing lawsuits under the law. Enforcement authority would rest exclusively with the Illinois attorney general.
Violations could result in civil penalties of up to $3 million per offense.
The legislation also includes disclosure requirements for covered AI developers, including designated contacts and business information. Companies subject to the law would contribute fees intended to help fund administration and oversight.
Another provision creates whistleblower protections and reporting mechanisms for employees of AI companies.
If signed by Governor Pritzker, Illinois would potentially establish a model that other states, and possibly federal lawmakers, may look to as AI regulation continues to evolve.
