New York opened an online portal Tuesday where the largest developers of advanced artificial intelligence must register with the state by Jan. 1, Gov. Kathy Hochul's office announced. The step turns the state's RAISE Act from a statute into a working compliance regime, overseen by the Department of Financial Services, that gives leading AI labs a fixed sign-up date and a 72-hour deadline to report serious safety incidents.

What Happened

The registration system, called the RAISE Portal, was launched by the Office of Digital Innovation, Governance, Integrity and Trust, known as DIGIT, a new unit inside the Department of Financial Services. DIGIT also published a website setting out what developers must do, according to the governor's release and a matching notice from the department.

The release says the law reaches all frontier AI developers, meaning companies that build the most capable general-purpose models, operating in New York, with extra duties for the largest. Large frontier developers must register by Jan. 1, 2027, and from January must post their safety frameworks, the written plans describing how they test for and limit dangerous capabilities, on their websites. All frontier developers must report critical safety incidents within 72 hours and publish transparency reports on their models.

RelatedOpenAI, Anthropic and Microsoft Back Paris AI Safety Call; U.S. and UK Not Among Signers

TechCrunch reported that the "large" tier covers companies with more than $500 million in annual revenue, including OpenAI and Anthropic. Penalties reach $1 million for a first violation and $3 million for later ones, and the state can act against companies that fail to register, TechCrunch reported; a spokesperson declined to give details.

Hochul also launched the New York Center for AI Responsibility and Research at Binghamton University, part of the State University of New York system, to test advanced AI systems independently, conduct safety research and train specialists. Greg Morrisett of Cornell Tech and Jeannette Wing of Columbia University are interim co-directors, and more than 80 researchers from New York institutions will take part. The center is funded by a $30 million gift from donors led by Bloomberg LP co-founder Tom Secunda, a $25 million SUNY research investment and state operating money, and it draws computing power from Empire AI, a $500 million public-private initiative. At the state's request, it will work with DIGIT to review developers' published safety frameworks and recommend stronger requirements.

"The stakes are too high to simply take AI companies at their word," Hochul said in the release.

Why It Matters

The 72-hour clock is the most demanding element for companies. California's SB 53, in force since January, gives frontier labs 15 days to report critical incidents, and Illinois' SB 315, which takes effect in 2027, adds independent compliance audits, TechCrunch reported. A lab operating nationally will have to build incident processes fast enough for New York's deadline.

Placing enforcement with the state's financial regulator, and pairing it with a state-backed lab that can test models, gives New York its own technical means to question developers' safety claims rather than relying on what the companies publish, the point Hochul made in launching both.

For companies that build on these models, the rules offer new information: published safety frameworks and transparency reports can feed vendor reviews and contract terms.

Context

New York's rules arrive while Washington remains split over whether federal law should override state AI statutes, an approach known as preemption. The Primary Wire reported in October, citing Reuters, that a political group funded by OpenAI President Greg Brockman backs candidates who favor preemption, while an Anthropic-funded group opposes it. The state has also moved on narrower AI rules, including a law on chatbot features for minors that The Primary Wire reported in October.

What to Watch

  • Which companies register before the Jan. 1 deadline, and whether DFS names or penalizes those that do not.
  • The first safety frameworks posted in January and the testing center's recommendations on them.
  • Whether Congress takes up preemption in a way that would override New York's requirements. Hochul said she will look for further ways to strengthen the state's AI safety laws in the coming months, according to the DFS notice.