Illinois has become the third state to mandate transparency and accountability measures for advanced AI systems, passing a landmark bill that requires the largest AI developers to publish detailed frameworks explaining how they assess model capabilities and safety risks. The legislation, Senate Bill 315, passed the Illinois House unanimously on May 28 and now heads to Governor JB Pritzker for signature, following similar laws in New York and California that lawmakers hope will eventually push the federal government to act.

The bill targets the most powerful AI systems developed by the largest companies, using specific thresholds to define scope: developers with at least $500 million in revenue and systems meeting a massive computing measurement standard must comply. Both OpenAI and Anthropic, two of the world's leading AI companies, supported the legislation throughout its development, signaling industry recognition that some form of accountability is inevitable.

What Would Illinois's AI Accountability Law Actually Require?

The bill imposes several concrete obligations on covered AI developers. Companies must create and publish transparency frameworks that explain how they apply industry standards, measure model capabilities, assess the chance of catastrophic risk, and identify and respond to safety incidents. A particularly contentious provision requires developers to employ third-party auditors to verify compliance with these frameworks, ensuring that companies cannot simply self-certify their own safety practices.

The legislation also establishes enforcement mechanisms with real teeth. The Illinois attorney general would have exclusive authority to enforce civil penalties up to $3 million per violation, giving state regulators a concrete tool to hold companies accountable. Large frontier AI developers must file disclosure statements with primary contacts and places of business and pay proportional fees to cover the cost of administering the law.

How to Understand the Third-Party Audit Requirement

• What Gets Audited: Third-party auditors verify that AI developers are following their own published transparency frameworks and properly assessing model capabilities and safety risks throughout development and deployment.
• Who Can Audit: The bill clarifies that qualified auditors include small boutique firms specializing in AI safety as well as large international accounting firms with the technical capabilities to perform meaningful assessments.
• Proprietary Protection: Amendments to the bill added protocols for protecting proprietary information during audits, addressing industry concerns that transparency requirements could expose trade secrets.
• Enforcement Timeline: The effective date was extended from 2027 to 2028, giving companies time to develop audit-ready frameworks before compliance obligations begin.

The third-party audit requirement proved controversial during the legislative process. TechNet, a coalition of technology executives, expressed concern that Illinois would be requiring companies to make subjective safety determinations without established national standards or clear regulatory guardrails. However, advocates like Secure AI, a nonprofit pushing for meaningful AI regulation, supported the requirement as essential for genuine accountability.

"This legislation enacts critical protections against the most catastrophic risks that advanced AI systems pose to public safety. Artificial intelligence is among the most significant technological developments of modern time. It has the potential to drastically improve the quality of life of people throughout the world, but only if deployed and developed responsibly," said Rep. Daniel Didech, the House sponsor of the bill.

Rep. Daniel Didech, D-Buffalo Grove, Illinois House of Representatives

House sponsor Rep. Daniel Didech explained that his office consulted with House legal staff and conducted research to identify companies capable of performing these audits. He noted that a developing ecosystem of boutique firms already exists alongside large international accounting firms with the necessary technical capabilities, making the requirement feasible.

Why Are States Moving Faster Than the Federal Government?

Illinois's action reflects a broader pattern of state-level AI regulation outpacing federal efforts. Legislators explicitly modeled the bill after 2025 laws in New York and California, hoping to advance what they describe as a "de facto" national standard in the absence of federal action. Senate sponsor Sen. Mary Edly-Allen compared the current AI landscape to the "wild, wild West," arguing that lawmakers cannot repeat the mistake they made with social media regulation, which remained minimal until recently.

The timing is significant. At the federal level, President Trump postponed a planned executive order that would have created a voluntary government testing process for advanced AI models, citing concerns that such oversight could slow innovation and U.S. competitiveness with China. This decision left a regulatory vacuum that states are now filling with their own mandatory frameworks.

The bill passed the Illinois House with overwhelming support: 110 votes in favor and zero opposed. The Senate had previously approved it 52 to 5, demonstrating rare bipartisan consensus on the need for AI accountability measures.

"As these models grow more powerful, this kind of enforceable accountability matters more than ever. Illinois lawmakers have set a new standard, and we hope other states and the federal government build on their dedication to AI safety," stated Cesar Fernandez, Anthropic's head of U.S., state and local government relations.

Cesar Fernandez, Head of U.S., State and Local Government Relations, Anthropic

Anthropic's public support for the bill is noteworthy given the company's involvement in advanced AI safety research. The company's backing suggests that leading AI developers increasingly view transparency and accountability requirements not as obstacles but as necessary components of responsible development.

What Happens Next in AI Regulation?

Illinois's passage of SB 315 adds momentum to a growing movement for state-level AI accountability. With New York and California already having enacted similar laws, and Illinois now joining them, other states are likely to follow. The legislation also includes a provision clarifying that the bill does not create a pathway for private citizens to sue, addressing concerns from some stakeholders while preserving the attorney general's enforcement authority.

Meanwhile, global AI governance continues to evolve. The European Union recently agreed to amend its AI Act by delaying compliance deadlines for high-risk systems, with biometric, education, employment, law enforcement, and critical infrastructure applications now facing a December 2, 2027 deadline, while AI systems embedded in products face an August 2, 2028 deadline. These international developments suggest that AI governance is becoming a coordinated, multi-jurisdictional effort rather than a purely national concern.

The broader challenge remains translating high-level governance principles into practical, enforceable mechanisms. Academic research has highlighted a persistent "principles-to-practice gap" in AI governance, where governments and organizations articulate values like fairness, transparency, and accountability but struggle to operationalize them in real-world settings. Illinois's requirement for published frameworks and third-party audits represents one concrete approach to bridging that gap.

As Illinois's bill heads to the governor's desk, it signals that states are willing to act decisively on AI accountability even as federal policymakers remain divided. Whether other states follow Illinois's lead, and whether the federal government eventually adopts similar standards, will shape how AI companies operate in the United States for years to come.