OpenAI, Anthropic face new oversight as New York’s AI safety law takes effect

New York opened a registration portal on Tuesday for the largest AI developers, kicking off enforcement of a state safety law that will pull OpenAI, Anthropic, and other frontier labs into ongoing state oversight beginning in 2027.

Under the RAISE Act, which takes effect at the start of the year, developers with more than $500 million in annual revenue must register by January 1, 2027, publish safety frameworks on their websites, issue public model transparency reports, and report critical safety incidents within 72 hours. That reporting clock is markedly tighter than the 15-day requirement under California's SB 53, which took effect in January.

The duties fall to the newly formed Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT), which will implement the act and oversee large frontier labs. Penalties run as high as $1 million for a single violation and $3 million for subsequent ones, and the state can also bring enforcement actions against companies that fail to register, though a spokesperson declined to give specifics on what those would involve.

Governor Kathy Hochul cast the law as a brake on industry boosterism at a Tuesday briefing, saying tech leaders declare AI's arrival inevitable and treat any harm as acceptable in pursuit of the race to "the AI cliff." She also launched the New York Center for AI Responsibility and Research at SUNY Binghamton, an independent body that will evaluate the safety, security, and operation of advanced AI systems and partner with DIGIT to review companies' safety frameworks. It builds on her January announcement backed by $55 million in donations and capital, and she described the whole apparatus as "a national model for smart, sensible AI regulation."

New York is stacking requirements across categories: a law in force since last year obliges AI companion operators to disclose that users are talking to a bot and to make reasonable efforts to detect suicidal ideation and self-harm, while a separate chatbot safety law for minors signed in October restricts deceptive human-like interactions, sycophantic or manipulative responses, and content encouraging suicide or disordered eating, effective June 1, 2027.

Illinois' SB 315 adds independent safety compliance audits when it starts in 2027, alongside California's already-active rules. Open questions for the year ahead: how DIGIT will exercise its enforcement discretion and spell out penalties for missed registrations, whether the $500 million revenue line captures the intended labs, and how companies will reconcile conflicting incident-reporting timelines across states.