Gov. JB Pritzker on Monday signed artificial intelligence legislation modeled on similar bills in California and New York, furthering a push for a state-driven national framework to replace federal regulations.
“Congress and the president should pass similar legislation, but they have been reluctant so far because many are beholden to special interests that profit from an unregulated industry,” Pritzker said before signing the bill. “We can work together to build thoughtful guardrails in a way that benefits the industry and the public, or we can allow a small number of actors to evade responsibility and shift the costs and damages to ordinary people. Illinois has chosen our path.”
Senate Bill 315Also known as the AI Safeguards Act, it increases transparency and accountability requirements for the largest artificial intelligence models, which generate more than $500 million in annual revenue and use massive amounts of computing power to train.
The bill reflects the situation in California SB-53 and New York Responsible Artificial Intelligence Safety and Education Actboth signed by the end of 2025. It sets new reporting standards for the potential for AI models to be used for large-scale harm, such as helping users create chemical, biological, or nuclear weapons or conduct cyberattacks.
Senate sponsor Sen. Mary Edley Allen, D-Libertyville, said states urgently need to take steps to prevent these potential harms.
“We are not willing to wait for Congress to act,” Edley-Allen said. “There’s an old saying: Give a man a fish and you feed him for a day. Give a man a fish and you feed him for a lifetime. But teach an AI how to fish and it might empty an entire river trying to figure out how to do that.”
Although these three states account for only about 20% of the nation’s population, lawmakers estimate they represent about 40% of the U.S. AI market, effectively setting the de facto national standard.
new guardrail
The new law requires model developers to publish an artificial intelligence framework outlining how developers can identify and assess “catastrophic risk,” defined as the likelihood of an event that could result in the death or serious injury of more than 50 people or more than $1 million in property damage.
Developers are also required to report any incidents that may cause harm to the state within 72 hours of becoming aware of the incident, or within 24 hours if the incident poses an imminent risk of death or serious bodily injury.
The bill’s House sponsor, Democratic Rep. Daniel Didech of Buffalo Grove, said the harms regulated by the bill are not theoretical.
“We’ve seen the first mass shootings powered by artificial intelligence. We’ve seen artificial intelligence systems being used to attack municipal water and sewer facilities,” Didtchi said.
He also mentioned an example Anthropic The company calls it the Myth model Cyber weapons are too powerful Release to the public. Anthropic supports the Illinois bill and sent representatives to Monday’s signing ceremony.
“Every transformative technology in our history, from cars to electricity to air travel, has brought huge benefits but also real risks, and in each case the government response has been not to ban the technology or to take a hands-off approach, but to put in place safeguards to give people confidence that these technologies are safe,” Diedge said.
Illinois’ version is similar in most respects to standards set by New York and California, adding the nation’s first mandatory annual third-party audit requirement; New York’s version only requires an independent audit when a developer is large enough to comply with the law.
The provision of third-party audits was a point of contention among some industry stakeholders during the conference debate, including TechNet, a coalition of technology executives across the industry.
“We remain concerned that Illinois will effectively require private actors to make highly subjective decisions requiring AI safety compliance without established national standards, certifications, or clear regulatory guardrails,” TechNet representative Ninia Linero said during the May 20 committee meeting.
OpenAI and Anthropic both supported the bill’s passage in the Illinois General Assembly, where it received broad bipartisan support in both chambers; only five Republican senators voted against unanimously In the House of Representatives.
While large developers are pushing for a federal framework rather than the patchwork of inconsistent state regulations they fear, Caitlin Niedemeyer of OpenAI Global Affairs told the Senate Artificial Intelligence and Social Media Committee in April that OpenAI is open to a coordinated, state-driven approach.
“While we’re very clear that the federal government is still well-positioned to lead on border security because it has the resources, the expertise and the institutions, we actually also feel strongly that Illinois, California and New York can all really lead on moving forward with a unified framework that we believe will absolutely help create a de facto national direction of travel,” Niedermeier told the committee.
What is the future of artificial intelligence?
Companies that violate this provision are subject to civil penalties from the Attorney General’s Office, which can range up to $1 million for a first violation and up to $3 million for subsequent violations.
But lawmakers and advocates say they hope to continue research on the topic of artificial intelligence in the future. For example, Diedchi believes health care and education may be frontier areas that require further assessment of the public safety risks of AI.
Scott Wisor, director of policy at Secure AI, was one of the advocates who helped craft the Illinois bill. He said more external assessment of the risks posed by the model and judgment on when it would be ready for release would be the next step to further improve transparency and accountability.
“Right now, the assessment in the bill is, are you complying with your security framework? Because let’s say you have a security framework that’s like, ‘We’re going to do A, B, C and D,’ and you do it, and the assessors confirm that, but it’s still a risky thing to implement in the world,” Wisor said.
“So, it’s a big step forward, but I think there’s more we can do,” Visor said.
Illinois’ law will take effect on January 1, 2028.
Copyright 2026 The Associated Press. all rights reserved. This material may not be published, broadcast, rewritten or redistributed.
theme
Illinois data-driven artificial intelligence legislation for insurtech
