New York's Responsible AI Safety and Education (RAISE) Act, effective January 1, 2027, closely mirrors California's SB-53 framework, requiring AI developers to publish safety testing and risk mitigation approaches and report critical safety incidents to state officials. Both laws apply to models trained on more than 10^26 FLOPS from companies with over $500 million annual revenue. Instead of fragmenting into conflicting state rules, the two largest states are converging on identical transparency and reporting standards, reducing compliance complexity for developers. For financial institutions, this signals that US frontier AI regulation may stabilize around a predictable "trust but verify" model rather than balkanize, lowering long-term policy uncertainty for AI vendors, though implementation and enforcement capacity remain untested at the state level.
320 other tracked developments touching the same claims.