Source: Jones Walker AI Law Blog

The Jones Walker AI Law Blog provides an in-depth analysis of New York’s Responsible AI Safety and Education (RAISE) Act, signed into law on December 19, 2025. This legislation specifically targets "Large Developers" of "Frontier Models"—defined as entities that have spent over $100 million in computing costs or trained models exceeding 10^26 computational operations. These developers are now required to maintain and publicly disclose comprehensive safety and security protocols. Furthermore, they must conduct annual reviews and report any safety incident that creates a risk of "critical harm" to the state’s Attorney General within 72 hours of determination.

The act also mandates a rigorous five-year document retention policy for all safety assessments and protocols, which must be made available for government audits. This state-level regulation enters a complex legal environment, following a federal executive order from President Trump aimed at creating a "minimally burdensome national standard" and preventing a patchwork of discordant state laws. Consequently, the RAISE Act may face constitutional challenges regarding interstate commerce and "compelled speech." Nevertheless, it represents a significant move by New York to hold the largest AI developers accountable for potential catastrophic risks, setting a high bar for transparency and corporate responsibility in the tech sector.