Illinois Governor JB Pritzker signed Senate Bill 315, the Artificial Intelligence Safety Measures Act (AISMA), into law on July 6, 2026, making Illinois the third state to require frontier model developer transparency and the first state to require third-party audits of artificial intelligence models. The law is structured around the same definitional core as its predecessors: it targets 'large frontier developers,' AI companies with annual gross revenues exceeding $500 million that train models using computing power greater than 10 to the 26th operations, a threshold that is likely to capture industry giants such as OpenAI, Anthropic, Google, Meta, and xAI. The three states collectively carry substantial weight: lawmakers estimate that California, New York, and Illinois account for roughly 40 percent of the U.S. AI market, thus effectively establishing a de facto national standard in lieu of federal regulations.
The compliance calendar is now compressed. California's Transparency in Frontier Artificial Intelligence Act (TFAIA), the first state law requiring standardized safety and transparency disclosures from frontier model developers, took effect January 1, 2026. New York Governor Kathy Hochul signed the final version of the Responsible AI Safety and Education Act (RAISE Act) on March 27, 2026, aiming to establish 'a strong and sensible standard for frontier AI safety, holding the biggest developers accountable for their safety and transparency protocols.' Both the RAISE Act and AISMA take effect January 1, 2027. Illinois goes a step further than either predecessor: the Illinois act requires developers to retain an independent third party to audit their compliance annually, with AI framework and independent audit provisions not becoming effective until January 1, 2028. The incident-reporting obligations are also stringent: AISMA imposes 72-hour reporting obligations for certain AI safety incidents and 24-hour reporting for incidents posing imminent risks of death or serious injury.
The practical significance of the three-state cluster is that geography no longer offers an escape. For AI models and other products that are not geographically confined, state transparency requirements can have national effects, because the resulting disclosures are publicly available. The United States now has three large states with very similar transparency laws governing frontier AI models, arguably reflecting an emerging de facto national standard in the absence of a federal statute. Enforcement authority in Illinois rests with the state attorney general: penalties reach up to $1 million for a first violation and up to $3 million for subsequent violations. The law was passed by overwhelming bipartisan margins, 110 to 0 in the House and 52 to 5 in the Senate, and was part of a broader package of bills the Democratic-controlled General Assembly pushed to fill what lawmakers view as a void left by federal inaction on AI regulation.
Skeptics and industry groups are not satisfied. During the General Assembly debate, the third-party audit provision was a point of contention for industry stakeholders, including TechNet, a coalition of tech executives, whose representative Ninia Linero said: 'We remain concerned that Illinois would effectively be requiring private actors to make highly subjective determinations requiring AI safety compliance without established national standards, certifications, or clear regulatory guardrails.' At the federal level, the law was enacted despite strong opposition by the Trump administration to any state laws that regulate AI developers. Some legal analysts note a plausible preemption scenario: organizations subject to the RAISE Act cannot assume federal preemption will eliminate compliance obligations before the effective date and should build required infrastructure now while monitoring constitutional challenges that may eventually invalidate requirements. These laws reflect a legislative focus on transparency and reporting rather than substantive restrictions on model development and deployment, a nuance that may or may not satisfy federal courts if challenges arrive.
Any frontier AI company that has not begun building its compliance infrastructure for January 1, 2027 is now simply late.