Illinois Governor Signs AI Safety Measures Act — First US State to Require Third-Party Audits of Frontier AI Models

What happened

Illinois Governor J.B. Pritzker on July 6, 2026 signed Senate Bill 315 (Artificial Intelligence Safety Measures Act). The law creates the strongest state-level AI safety regulatory framework in the US—the first to mandate independent audits of large frontier models.

Context and impact

The law targets "large frontier developers"—companies with at least $500M in annual revenue and massive computing power—meaning OpenAI, Anthropic, Google, and similar firms. It goes further than New York and California regulations by requiring annual third-party audits. Penalties: $1M for the first violation, $3M for each subsequent violation.

Details

• Safety plans: companies must publish and adhere to them
• First US law requiring annual independent audits of frontier AI models
• Critical incidents: must be reported within 72 hours; imminent risk of death/injury within 24 hours
• Whistleblower protections for employees raising AI safety concerns
• Confidential reporting channels for internal incident disclosure