Illinois Lawmakers Pass Bill Requiring Safety Disclosures on ‘Frontier’ AI Models

The law will set a new standard for regulating so-called “frontier” AI models. Once signed by Pritzker, Illinois will become the third state to establish a regulatory standard for frontier models, after New York’s RAISE Act and California’s Transparency in Frontier AI Act.

The Illinois law requires AI companies to publish and annually update a plan to address “catastrophic risks” from their models and mandates third-party audits of safety issues.

“This legislation enacts critical protections against the most catastrophic risks that advanced AI systems pose to public safety,” said state Rep. Daniel Didech, who sponsored the bill in the House, according to the Illinois News-Gazette. “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.”

Related: In an About-Face, Trump Administration Considers Vetting New AI Models

Sen. Mary Edly-Allen, the measure’s chief sponsor in the upper chamber, warned that lawmakers cannot take the same hands-off approach they initially took with social media.

“This is not about stopping innovation, but rather about balancing the great promise of AI with its potential harms,” said in introducing the bill earlier this month.

In addition to requiring developers to publish safety plans before deploying newly developed models, the law applies to “substantially modified” iterations of existing models. It also directs the Illinois Emergency Management Agency and Office of Homeland Security, in consultation with the Attorney General, to administer reporting mechanisms, issue guidance, and prepare annual reports.

The Illinois law comes even as the White House continues to push for a federal standard for AI that would preempt state laws. In March, the administration released a proposed federal legislative framework for AI that called for Congress to override state laws. “This framework can only succeed if it is applied uniformly across the United States,” the White House said in a statement on the framework. “A patchwork of conflicting state laws would undermine American innovation and our ability to lead in the global AI race.”

Nonetheless, both OpenAI and Anthropic supported the Illinois law. “As these models grow more powerful, this kind of enforceable accountability matters more than ever,” Cesar Fernandez, Anthropic’s head of U.S., state and local government relations, said in a statement Wednesday. “Illinois lawmakers have set a new standard, and we hope other states and the federal government build on their dedication to AI safety.”

Illinois may not be through with AI-related legislation. At least eight other such bills remain pending with three days left in the current legislative session. They include bills on data privacy, a ban on algorithmic rental price-fixing, and the use of AI in health care approvals, among other things. All had been approved in the Illinois House are awaiting action in the Senate.