Illinois Governor JB Pritzker signed the Artificial Intelligence Safety Measures Act on July 6, 2026, making Illinois the first state to require annual independent third-party audits of the country's largest AI companies. The law targets "frontier developers" — companies with more than $500 million in annual revenue that train the most powerful AI models. They now have to publish safety plans covering catastrophic risks like weapons of mass destruction or autonomous cyberattacks. They must also submit to outside audits every year and report critical safety incidents within 72 hours. It takes effect in January 2028.

The law exposes a split that runs straight through the tech industry: OpenAI and Anthropic, the two AI companies most directly regulated by it, both supported it. Google, Apple, and Amazon — through their trade groups — fought to kill it. And the Trump administration has a federal task force whose explicit job is to sue states that pass laws like this one.

---

Illinois Governor JB Pritzker says the state acted because Congress wouldn't.

Pritzker framed the law as filling a hole that Washington refuses to fill. "Congress and the president ought to be passing similar legislation," he said at the signing, "but they've so far been unwilling." State Sen. Mary Edly-Allen, the Senate sponsor, called the AI industry "the Wild Wild West" and made the state's position blunt: "We are not willing to wait for Congress to act."

State Rep. Daniel Didech, the House sponsor, cited "the first AI-inspired mass shooting" as a reason to act during floor debate. Illinois AG Kwame Raoul said the $3 million maximum fine is "perhaps not enough" deterrent for trillion-dollar companies. Republican House minority leader John Curran co-sponsored the bill, calling it "a bold first step."

Illinois, California, and New York together cover about 40% of the U.S. AI market. Pritzker and the bill's sponsors are betting that if the biggest states set consistent rules, those rules become the national floor — with or without Congress.

---

OpenAI and Anthropic both supported the law.

OpenAI policy director Caitlin Niedermeyer backed the bill throughout the legislative process. The company's position: the federal government is "well-positioned to lead," but a "coordinated state-driven approach" creates a "de facto national direction of travel" in the meantime. Anthropic called the law "an important step toward the accountability this technology demands."

Both companies already conduct internal safety reviews. Anthropic cited its own decision not to release its "Mythos" model publicly because of safety concerns. The Illinois law formalizes that kind of internal review into an externally audited, public-facing requirement.

Instead of waiting for a federal standard, OpenAI and Anthropic are backing state-by-state rules that effectively converge into one national framework. OpenAI and Anthropic already run safety programs, so the audit requirement adds outside accountability without changing what they actually do.

---

The platform companies say no qualified auditors exist yet.

Chamber of Progress — backed by Google, Apple, and Amazon — sent a letter opposing the bill the morning of the House vote. TechNet representative Ninia Linero called the audit requirement "highly subjective determinations" made "without established national standards, certifications, or clear regulatory guardrails." The Computer and Communications Industry Association said the state is requiring companies to comply with "a process for which the State has not yet established the necessary infrastructure, standards, or qualified evaluators."

Their core objection is simple: no qualified independent AI safety auditors exist in the U.S. yet. There is no licensing body, no certification framework, no established methodology. Chamber of Progress said companies would have to pass audits run by auditors whose qualifications nobody has defined — "all liability and no standards."

Google, Apple, and Amazon are not themselves the frontier model developers the law targets. They are platform companies whose products run on top of AI models and who would face downstream compliance questions.

---

Trump created a DOJ task force to challenge state AI laws.

In December 2025, Trump signed Executive Order 14365 — "Ensuring a National Policy Framework for Artificial Intelligence" — declaring state AI regulation a harmful "patchwork of 50 different regulatory regimes." The order created a DOJ AI Litigation Task Force with the sole purpose of challenging state AI laws in federal court on Commerce Clause, preemption, or other grounds. The order also makes states with "onerous AI laws" ineligible for federal broadband funding.

In April 2026, the DOJ put the task force to work: it joined xAI's lawsuit against Colorado's AI Act, the first time the federal government intervened against a state AI law. Colorado subsequently repealed the law and passed a narrower version that dropped its anti-discrimination requirements entirely. Illinois SB 315 reads differently — it targets catastrophic physical safety risks, not algorithmic discrimination, which was the Colorado law's core. Legal analysts say Illinois's focus on physical safety — rather than algorithmic discrimination — gives it stronger constitutional footing. But the task force exists, and Illinois just handed it a new target.

---

Where This Lands

Illinois passes first. The federal government fights states that try to regulate AI. The biggest AI companies support state oversight because they already do what the audits would formalize. Consistent state-level rules also give them more certainty than waiting for Congress. The platform companies behind them oppose the audits because no qualified auditors exist and the compliance burden falls on them without any standards to follow.

No one disputes that catastrophic AI risks are real. The disagreement is about who sets the rules and when — and whether mandating audits without established standards actually makes companies safer or just creates liability for a process nobody has defined. Illinois's law includes a federal override: if Congress passes equivalent rules, the state requirements drop away. Congress hasn't.

---

Sources

- Capitol News Illinois — Pritzker signs landmark AI regulation bill: https://capitolnewsillinois.com/news/pritzker-signs-landmark-ai-regulation-bill-that-aims-to-mitigate-risks/ - Chicago Sun-Times — Gov. Pritzker signs Illinois AI regulations into law: https://chicago.suntimes.com/politics/2026/07/06/ai-regulations-illinois-law-pritzker-signed - Governing — Illinois Moves to Become the First State to Mandate AI Safety Audits: https://www.governing.com/artificial-intelligence/illinois-moves-to-become-the-first-state-to-mandate-ai-safety-audits - Akerman LLP — Illinois SB 315: A State Strategy for Enduring National AI Safety Standards: https://www.akerman.com/en/perspectives/illinois-sb-315-a-state-strategy-for-enduring-national-ai-safety-standards.html - Gizmodo — Illinois Drops the Hammer on AI Companies: https://gizmodo.com/illinois-drops-the-hammer-on-ai-companies-2000781932 - ABC7 Chicago — Illinois governor signs AI bill into law: https://abc7chicago.com/post/illinois-governor-jb-pritzker-sign-ai-bill-law/19457902/ - White House — Executive Order 14365: https://www.whitehouse.gov/presidential-actions/2025/12/eliminating-state-law-obstruction-of-national-artificial-intelligence-policy/ - Jenner & Block — DOJ Joins xAI in Lawsuit Challenging Colorado AI Act: https://www.jenner.com/en/news-insights/client-alerts/doj-joins-xai-in-lawsuit-challenging-colorado-ai-act - DOJ — Justice Department Intervenes in xAI lawsuit: https://www.justice.gov/opa/pr/justice-department-intervenes-xai-lawsuit-challenging-colorados-algorithmic-discrimination - NPR Illinois — Pritzker signs landmark AI regulation bill: https://www.nprillinois.org/illinois/2026-07-06/pritzker-signs-landmark-ai-regulation-bill-that-aims-to-mitigate-risks