Bill regulating powerful AI models advances in Illinois as advocates say it’s only the first step
The Illinois Senate advanced Senate Bill 315, requiring transparency and safety protocols for large AI developers like Meta, OpenAI, and Anthropic, with a focus on catastrophic risk reporting. The bill, modeled after New York and California laws, passed 52-5 despite federal resistance, as advocates argue it sets a foundational standard for national AI governance.
Illinois lawmakers took a major step toward regulating powerful AI models with the passage of Senate Bill 315, which requires developers like Meta, OpenAI, and Anthropic to adopt transparency frameworks, conduct third-party audits, and disclose catastrophic risk capabilities. The bill, introduced by State Sen. Mary Edly-Allen, targets companies with revenues exceeding $500 million, mirroring recent legislation in California and New York. It passed the Illinois Senate 52-5 on May 13, despite opposition from federal officials, including President Donald Trump, who previously discouraged state-level AI regulations. The legislation defines 'frontier' AI companies as those at the cutting edge of technology, including firms whose models have raised concerns about misuse, such as Anthropic’s Mythos, which the company refused to release due to safety risks. Testimony from Secure AI’s policy director, Scott Wisor, highlighted that some AI models could assist in developing biological weapons, underscoring the need for stricter oversight. OpenAI and Anthropic supported the bill, emphasizing the importance of national standards focused on the most advanced and potentially harmful models. Critics, including Jonathan Iwry of the Wharton Accountable AI Lab, noted that the bill’s requirements—such as transparency, testing, and auditing—are modest but essential given the unpredictable nature of AI systems. They argue that while these safeguards are foundational, they fall short of full accountability. The bill’s passage follows Trump’s recent proposal for a second AI-related executive order, though state lawmakers have pushed forward independently due to perceived federal inaction. The legislation is part of an eight-bill package aimed at establishing a de facto national AI standard, with industry lobbyists urging consistency across states. Advocates stress that Illinois’s move is just the first step in a broader conversation about regulating AI, as technology continues to evolve beyond current safety measures. The bill now awaits further legislative action before potentially becoming law.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.