AI Frontier Institute

AI PRODUCT LIABILITY ACT

View primary source →

Illinois SB3502, introduced February 5, 2026 by Sen. Rachel Ventura, creates the Artificial Intelligence Design Requirements Act (Section 1). The bill establishes a product liability framework specifically for high-impact and generative AI systems. Under Section 15, AI developers can be held liable for: (1) defective design, if the developer failed to exercise reasonable care and a feasible alternative design existed; (2) failure to provide adequate instructions or warnings about known dangers; and (3) failure to conform to an express warranty on which the plaintiff reasonably relied. Under Section 20, a deployer of an AI system is treated as a developer and faces the same liability if the deployer makes a material and substantial change to the product or intentionally misuses it contrary to the express warranty, causing harm. Small deployers with fewer than 20 employees or fewer than 10,000 users are excluded from the definition of 'deployer.' Section 25 sets out enforcement rules, including a comparative negligence standard under which a plaintiff's recovery is reduced but not barred by their own fault, and joint and several liability for developers and deployers. Courts must recognize a rebuttable presumption of non-defectiveness if a deployer has conducted documented testing and auditing consistent with the NIST AI Risk Management Framework, mitigated foreseeable risks, disclosed risks, maintained an AI data sheet available to the Attorney General, and implemented a documented risk management policy. Products used strictly for peer-reviewed scientific research are exempt. The Act supplements, rather than supersedes, existing common law and statutory product liability claims.

← Back to the tracker