Illinois AI Safety Act: Another Piece of the U.S. Regulatory Puzzle

The US continues to expand its AI regulatory landscape, but once again, it is happening primarily at the state level rather than through a comprehensive federal framework.

Illinois has now joined the growing list of states introducing dedicated AI legislation. Governor signed SB315, the Artificial Intelligence Safety Measures Act, which has been described as one of the most comprehensive state AI safety laws in the country. The Act focuses on transparency, accountability and governance for the largest frontier AI developers, while attempting to balance innovation with public safety. It will take effect on 1 January 2027.

Photo: macrovector na Magnific

When discussing AI regulation in the US, it is no longer enough to analyse federal initiatives alone. Organizations increasingly need to navigate a complex legal patchwork consisting of federal laws, state AI regulations, state privacy acts, sector-specific healthcare requirements, FTC enforcement, consumer protection laws and numerous industry regulations. This growing regulatory fragmentation significantly increases governance and compliance challenges.

The Illinois AI Safety Measures Act introduces requirements for the largest frontier AI developers, including:

➡️ publication of AI safety frameworks,

➡️ identification and mitigation of catastrophic AI risks,

➡️ transparency reporting before deploying frontier models,

➡️ reporting of significant AI safety incidents,

➡️ documented internal governance processes,

➡️ cybersecurity protections for unreleased model weights,

➡️ regular updates of AI safety frameworks,

➡️ mandatory independent third party safety audits,

➡️ whistleblower protections for employees reporting AI safety concerns.

AI appears to be following the path of a relatively fragmented area before evolving into a broad ecosystem of international, federal, state and sector-specific requirements. The consequence is that organizations can no longer build governance programs around a single regulation. Instead, they need integrated AI governance capable of mapping overlapping obligations across multiple jurisdictions and regulatory frameworks. Illinois will not be the last state introducing dedicated AI legislation.

Until the US adopts a comprehensive federal AI framework, we should expect this regulatory mosaic to continue expanding, adding yet another layer to the already complex landscape.

Link

Author: Sebastian Burgemejster