Sen. Mary Edly-Allen (D) and Rep. Daniel Didech (D), with bipartisan cosponsors including Sen. Sue Rezin (R) and House Republican Leader Tony McCombie · 2026

Illinois SB 315 — Artificial Intelligence Safety Measures Act (Public Act 104-0538)

Illinois SB 315

Illinois's frontier AI law, signed July 6, 2026 after passing the House 110–0, and effective January 1, 2027. It takes the SB 53 transparency model and adds verification. Large frontier developers (frontier models above 10^26 operations, developer revenue over $500M) must write, implement, publish, and annually update a frontier AI framework covering catastrophic-risk thresholds and assessment, mitigations, third-party evaluation, cybersecurity for model weights, internal governance, and — explicitly — catastrophic risk from internal use of their models, including models circumventing oversight. Transparency reports are due before or with each new or substantially modified deployment. From 2028, large developers must retain an independent third party for an annual compliance audit, publish a summary within 30 days, and send summaries of internal-use risk assessments to the state every three months. Critical safety incidents go to the Illinois Emergency Management Agency and the Attorney General within 72 hours, or to law enforcement within 24 hours when lives are at risk. Developers must register and pay fees to operate in the state. Penalties, enforced exclusively by the Attorney General, run to $1M for a first violation and $3M for subsequent ones. An interoperability clause lets the state designate federal regimes with equivalent incident reporting and independent audits as satisfying the Act.

Key Provisions

Regulatory Philosophy

SB 53 with an auditor. Illinois keeps California's disclosure architecture — frameworks, transparency reports, incident reporting, whistleblowers — and closes its largest gap by requiring someone outside the company to check compliance every year. It goes further than either California or New York on internal use, the risk that the most capable models are run inside the lab long before, or without, public release. And it is written to be absorbed: the interoperability clause invites a federal regime to satisfy it, anticipating national legislation rather than resisting it.

Where the burden falls

FrontierPrimaryAppsNoneHyperscalersNoneChipsNone
Base assessed
Frontier developers above 10^26 operations with more than $500M in revenue
Why it lands there
The same base as SB 53 and the RAISE Act, so the three state laws now land on the same handful of frontier developers with converging obligations. The registration fee is a direct charge to that layer for the cost of overseeing it. Nothing reaches deployers, the clouds, or the chip layer.
Compare all proposals by layer →

In contrast

Illinois SB 315 vs. California SB 53: the same model, with an auditor

Illinois copied California's structure and most of its definitions — the 10^26 threshold, the $500M large-developer line, frameworks, transparency reports, incident reporting, whistleblowers — and then added what SB 53 lacks. An independent auditor checks compliance every year. Internal use of frontier models is assessed and reported quarterly. Developers register and pay for their own oversight. Incident reporting is faster, at 72 hours rather than 15 days. California has since moved in the same direction with SB 813's verification framework, and Illinois's interoperability clause points beyond both: it is written to stand down in favor of an equivalent federal regime.

Compare with CA SB 53→

Strengths

Derived from the proposal’s own policy documents

  • +Adds independent annual audits to the transparency model, so published frameworks are checked against practice rather than taken on trust
  • +The most explicit state treatment of internal-use risk, including quarterly reporting to the state on models used inside the lab
  • +Unanimous House passage and bipartisan Senate cosponsors, including the House Republican Leader — the broadest coalition behind any state frontier law
  • +Shares its definitions of frontier model and catastrophic risk with California and the FRONTIER Act, reducing the patchwork that preemption advocates cite
  • +The interoperability clause offers Congress a path to harmonize without preemption: pass an equivalent federal regime and Illinois steps aside on its own terms

Weaknesses

From the perspective of political opposition

  • −Audits begin in 2028, more than a year after the law takes effect, and verify compliance with the developer's own framework rather than the adequacy of that framework
  • −Like SB 53 and RAISE, there is no gate: a model can ship before any audit finds a problem, and the state has no authority to stop it
  • −Penalties capped at $1M and $3M per violation are small relative to frontier developer revenue
  • −Administered by an emergency-management agency without frontier-AI technical capacity, which must stand up reporting systems and audit oversight from scratch
  • −A direct target of federal preemption: the FRONTIER Act and the Thune–Klobuchar draft would both displace its transparency, audit, and incident provisions

Position on Analytical Frameworks

Enforcement Mechanism vs. Regulatory Scope

Prevention vs. Liability & Regulatory Authority

Innovation Priority vs. Worker Protection

Pre-deployment Obligations vs. Federal Preemption

← Back to all proposals