Illinois to Enact AI Safety Law in 2027... Catastrophic Risks Must Be Reported Within 72 Hours

The U.S. state of Illinois has finalized safety legislation targeting major AI companies. Under the 'Artificial Intelligence Safety Measures Act,' effective January 1, 2027, companies with annual revenues exceeding $500 million must submit a 'transparency report' detailing risk assessments and mitigation plans—and undergo independent audits—when developing or using frontier models. They are also required to report any major safety incidents within 72 hours.

The law specifically defines 'catastrophic risk.' This refers to significant risks where the development, storage, use, or deployment of a frontier model is predicted to substantially contribute, in a single incident, to the deaths or serious injuries of 50 or more individuals, or property damage exceeding $1 million. In cases involving an imminent risk of death or serious injury, notification must be made within 24 hours.

In the event of violations, the Illinois Attorney General may impose civil penalties. Initial violations can result in fines of up to $1 million, while subsequent violations can lead to fines of up to $3 million.

Teri Ross, director of Illinois Legal Aid Online, explained that the law "requires companies to develop plans and undergo independent audits of their systems." Ross added, "Technology is changing so rapidly" and noted that "in some ways, it is hard for the law to keep up."