New York has opened registration under its RAISE Act, the first state AI safety law of its kind, and large frontier AI developers must sign up by January 1, 2027. Governor Kathy Hochul announced the portal on November 10, alongside a new research centre to test advanced AI systems.
What the Law Demands
The law applies to large frontier developers, defined as companies with more than $500 million in annual revenue. Those firms must publish their AI safety frameworks on their websites, report critical safety incidents within 72 hours, and issue public reports on the transparency of their models.
Enforcement sits with a new state office, the Office of Digital Innovation, Governance, Integrity, and Trust, known as DIGIT. It operates inside the Department of Financial Services and will review the safety frameworks that companies publish, with the state attorney general bringing civil actions.
Penalties run to $1 million for a first violation and $3 million for later ones. The state can also act against companies that fail to register, although a spokesperson would not say how those cases would be handled. The clock is now running toward the January deadline.
A Research Centre to Check the Labs’ Homework
Hochul also launched the New York Center for AI Responsibility and Research at the State University of New York’s Binghamton campus. More than 80 researchers will use Empire AI computing power to test advanced systems and judge whether their safety claims hold up.
Greg Morrisett of Cornell Tech and Jeannette Wing of Columbia University will lead the centre as interim co-directors. The centre was first announced in January with $55 million in donations and capital, and it will review company safety frameworks alongside DIGIT and suggest ways to strengthen them.
Hochul framed the move as a direct answer to federal inaction. ‘While the federal government fails to act, New York is leading the way on AI safety and holding big tech accountable,’ she said. ‘The stakes are too high to simply take AI companies at their word.’
Why It Matters
New York is the first state to turn AI safety promises into a registration system with deadlines and fines. OpenAI and Anthropic are the most prominent names in scope, and their published safety frameworks will now face review by state-appointed researchers rather than sitting unread on corporate websites.
The 72-hour incident rule is the sharpest in the country, far tighter than California’s 15-day window under SB 53. That speed matters because a dangerous model behaviour discovered on a Friday would reach regulators before the weekend is out, not two weeks later.
Our Read
Our read: the real test is enforcement, not paperwork. A spokesperson would not explain how the state will pursue companies that ignore the portal, and no investigator headcount or audit schedule has been disclosed. The law’s bite will be judged by its first penalty, not its first registration.
The timing also matters. With the Anthropic IPO pricing due on Wednesday and Washington debating liability relief for cyber incident reports, New York has chosen the opposite direction: more disclosure, faster reporting, and real money at stake for labs that stay silent.
Background
New York passed the RAISE Act last year and amended it in March to align its thresholds with California’s frontier AI law. The state already enforces a companion-bot disclosure rule, and a separate chatbot safety law for minors signed in October takes effect on June 1, 2027.
Illinois will follow with its own frontier AI law at the start of 2027, adding independent safety compliance audits to a similar set of requirements. Taken together, the state laws are building a patchwork of oversight while Congress has yet to pass a federal AI safety statute.
Source: TechCrunch’s report and Governor Hochul’s announcement


0 Comments