Illinois AI Safety Measures Act: What SB 315 Requires of Frontier AI Developers

Illinois AI Safety Measures Act: What SB 315 Requires of Frontier AI Developers

On July 6, 2026, Governor JB Pritzker signed the Artificial Intelligence Safety Measures Act (SB 315) into law, making Illinois the first state to require annual third-party safety audits of frontier AI systems, mandatory catastrophic-risk plans, and 72-hour incident reporting, effective January 1, 2027.

Information last verified on July 8, 2026. This is a developing story; we update it as the record changes.

Status: Signed July 6, 2026; effective January 1, 2027.

Jurisdiction scope: The Artificial Intelligence Safety Measures Act applies statewide in Illinois to AI companies that cross the compute threshold for a frontier model (roughly 10^26 computing operations). Its incident-reporting and whistleblower duties reach any such "frontier developer" regardless of revenue; its public risk-framework and annual-audit duties reach only "large frontier developers," those that also generate more than $500 million in annual revenue. It does not create new obligations for small AI vendors, ordinary businesses that merely use AI tools, or consumers, and it does not replace Illinois' existing AI employment discrimination law.

What Happened

Gov. JB Pritzker signed SB 315, the Artificial Intelligence Safety Measures Act, on July 6, 2026, after the bill cleared the Illinois General Assembly with bipartisan support. The bill was sponsored by state Sen. Mary Edly-Allen (D-Grayslake) and Rep. Daniel Didech, according to Capitol News Illinois and the governor's office.

At the signing, Pritzker described the measure as the strongest of its kind in the country: "I'm very, very glad to be with you today to enact the best artificial intelligence safety and accountability bill in the entire nation," he said, remarks reported by Capitol News Illinois and corroborated by multiple outlets covering the signing event. He also argued for federal action, saying Congress and the president "ought to be passing similar legislation, but they've so far been unwilling, because many are captive to special interests that profit from the industry having no regulation."

The Act, per the Chicago Sun-Times' account of the signed legislation, requires "large frontier developers" (those generating at least $500 million in yearly revenue from frontier-scale models) to "create, publish and adhere to safety plans to stave off dangerous uses of their rapidly advancing platforms." Those plans must lay out how a developer will mitigate "catastrophic risk," which the law ties to conduct that could "materially contribute to the death of, or serious injury to, more than 50 people," or cause more than $1 million in property damage.

The law's incident-reporting clock starts once a developer identifies a qualifying event: 72 hours for a general critical safety incident, or 24 hours if the incident poses an imminent risk of death or serious physical injury. The Act also includes whistleblower protections for employees who raise safety concerns internally or to regulators.

The governor's official newsroom page and the Illinois General Assembly's bill-status page each returned a 403 error to direct fetching during this research pass; the facts above were independently corroborated through the Chicago Sun-Times, Capitol News Illinois, and WGN-TV, which describe the same audit, reporting-window, whistleblower, and penalty provisions.

What the Law Actually Says

SB 315 does not regulate AI broadly. It targets only "frontier developers," entities that train, or begin training, a frontier model using computing power above roughly 10^26 operations, and reserves its toughest obligations for "large frontier developers," those that also clear $500 million in annual gross revenue. That two-part gate means the law reaches a small number of the largest AI labs rather than every business that deploys an AI chatbot or analytics tool.

The law layers its obligations in two tiers. For "large frontier developers" specifically, it requires a public catastrophic-risk framework, a written plan describing how the developer identifies, assesses, and mitigates the risk that its models contribute to mass-casualty or mass-damage events, plus an audit requirement: Illinois is the first state to mandate that these frameworks be checked every year by an independent third-party auditor, rather than through the single one-time review that comparable 2026 laws in New York and California require. A third obligation, incident reporting, reaches more broadly: any frontier developer that crosses the compute threshold, not only large frontier developers, must report a critical safety incident to the state within 72 hours of discovery, shortened to 24 hours if there is an imminent risk of death or serious physical injury. The Act's whistleblower protections likewise extend to employees of any frontier developer, not only large ones.

The Act pairs those duties with an enforcement structure worth noting. There is no private right of action; an individual harmed by a frontier AI system cannot sue a developer directly under SB 315. Enforcement runs exclusively through the Illinois Attorney General's office, alongside a carve-out preserving employees' separate rights under the state's existing whistleblower law. Civil penalties for noncompliance, failing to audit, filing a false or misleading framework, or missing a reporting deadline, run up to $1 million for a first violation and up to $3 million for each one after that.

SB 315 sits alongside, rather than inside, Illinois' existing AI statute book, covered in full in our Illinois AI laws guide: the HB 3773 employment-discrimination amendment to the Illinois Human Rights Act, the 2020 AI Video Interview Act, and the state's deepfake and digital-likeness statutes. None of those laws touch frontier-model safety auditing, the gap SB 315 fills. Illinois' biometric privacy law (BIPA) separately regulates how companies, including AI vendors, collect biometric identifiers, and the state's AI meeting-recording rules separately govern AI notetaking tools used on calls. SB 315 amends neither regime; a frontier developer touching biometric data or call recording in Illinois still must satisfy those laws independently. It also has no direct effect on Illinois' two-party consent recording law, which governs security cameras and audio capture rather than AI model behavior.

Analysis: Why This Matters

The following is analysis from the Recording Law Editorial Team. SB 315's significance lies less in what it bans than in what it forces into the open. Frontier AI developers have generally set their own internal safety practices on their own schedule. SB 315 converts that voluntary practice into a statutory duty for the largest players, backed by an outside audit rather than self-certification, and a state-enforced reporting clock rather than a company's own disclosure timetable.

The compute gate matters. By reaching only frontier developers, and reserving its toughest framework-and-audit duties for the subset that also clear $500 million in annual revenue, the law avoids sweeping in the thousands of Illinois businesses that merely use AI tools, a distinction the state's separate employment-discrimination AI law (HB 3773) already handles. SB 315 should not be read as a general AI-use statute; it is a safety-and-transparency regime aimed at the handful of companies capable of training the largest foundation models.

The absence of a private right of action also cuts in a specific direction: individuals who believe a frontier AI system caused them harm cannot bring a claim under this statute. Enforcement, and the leverage of $1 million to $3 million penalties, sits entirely with the Attorney General's office, concentrating both litigation cost and enforcement discretion in one state office.

How This Affects You

If you work for, or contract with, a company that trains large-scale AI models, the practical trigger is the compute threshold, not simply "using AI." A frontier developer that crosses that compute threshold but stays under $500 million in annual revenue is not subject to SB 315's framework or audit duties, but it can still be on the hook for the Act's incident-reporting and whistleblower provisions. Businesses that neither train frontier models nor cross the compute threshold are not subject to SB 315 at all, though they may still be covered by Illinois' other AI, biometric, or employment statutes. Employees at a covered developer who raise a safety concern internally are protected from retaliation under the Act's whistleblower provisions, separate from any rights they already have under Illinois' general whistleblower law. Consumers cannot sue a developer directly under SB 315; a complaint about a frontier AI system's safety practices would need to go through the Illinois Attorney General's office rather than private litigation.

This is general legal information, not legal advice. It covers Illinois and reflects sources verified on July 8, 2026. Laws change and this story is developing; consult a lawyer licensed in your jurisdiction about your specific situation.

Related articles

• Illinois AI Laws: Full Guide to Employment, Deepfake, and Digital Likeness Statutes
• AI Laws by State: 2026 Regulatory Tracker
• Illinois AI Meeting Recording Laws
• Illinois Biometric Privacy Law (BIPA)
• Illinois Recording Laws: Security Cameras

Last updated: 2026-07-08. This is a developing story; details verified as of 2026-07-08.

Frequently Asked Questions

What is Illinois SB 315, the Artificial Intelligence Safety Measures Act?

SB 315 is a 2026 Illinois law, signed by Gov. JB Pritzker on July 6, 2026 and effective January 1, 2027, that requires 'large frontier developers' of advanced AI models to publish a catastrophic-risk framework, undergo annual independent third-party audits, and report critical safety incidents to the state.

Who does the Illinois AI Safety Measures Act actually cover?

Its framework and audit duties cover 'large frontier developers' only: entities that train a frontier AI model using more than roughly 10^26 computing operations and that also generate more than $500 million in annual gross revenue. Its incident-reporting and whistleblower duties reach any 'frontier developer' that crosses the compute threshold, regardless of revenue. Ordinary business users of AI tools that do not train frontier models are not covered.

How fast must an AI safety incident be reported under SB 315?

A covered developer must report a critical safety incident to the state within 72 hours of discovering it, or within 24 hours if the incident poses an imminent risk of death or serious physical injury.

Can I sue an AI company under the Illinois AI Safety Measures Act?

No. SB 315 does not create a private right of action. Enforcement is exclusive to the Illinois Attorney General's office, aside from separate claims an employee may bring under Illinois' existing whistleblower law.

What are the penalties for violating SB 315?

Civil penalties reach up to $1 million for a first violation and up to $3 million for each subsequent violation, enforced by the Illinois Attorney General.

When does the Illinois Artificial Intelligence Safety Measures Act take effect?

The law takes effect January 1, 2027, following its signing by Gov. Pritzker on July 6, 2026.

Does SB 315 protect employees who report AI safety concerns?

Yes. The Act includes whistleblower protections that bar a covered developer from retaliating against an employee who reasonably reports a safety concern or a suspected violation of the law.

How does Illinois' law compare to New York's and California's AI safety laws?

Illinois' law goes further than comparable 2026 measures in New York and California by requiring an independent third-party audit of a covered developer's compliance every year, rather than a single one-time review.