New York Governor Kathy Hochul announced on November 10, 2026 that the state has launched the New York Center for AI Responsibility and Research (NYCAIRR) at Binghamton University, the first state run program to independently test frontier AI systems for safety instead of relying on companies to check their own work.
What Hochul announced
The announcement, carried in a release from the New York Department of Financial Services (DFS), formally launches a center Hochul first proposed in her January 2026 State of the State address. “While the federal government fails to act, New York is leading the way on AI safety,” Hochul said in the release, adding that the state cannot simply take AI companies at their word.
NYCAIRR is built on a $30 million philanthropic gift led by Binghamton alumnus and Empire AI board chair Tom Secunda, paired with a $25 million initial research investment from the State University of New York (SUNY) system and operating capital from the state. Empire AI, Hochul’s $500 million public private AI computing partnership, will supply the computing power the center’s researchers need to run safety evaluations.
More than 80 scientists from New York colleges and research institutions are taking part, working in AI safety, security and computer science while remaining at their home institutions. Cornell Tech Dean Greg Morrisett and Columbia University research chief Jeannette Wing will serve as interim co-directors, setting the center’s scientific agenda and overseeing its independent evaluations.
What the center will actually test
NYCAIRR’s stated mandate covers three things: independent testing of frontier AI systems for safety and security risks, foundational safety research, and training a new generation of AI safety specialists. Its first recruiting priority, per the DFS release, is researchers working on formal verification, interpretability and alignment, hardware security and AI biosecurity.
At the state’s request, the center can also feed assessments into policy decisions. One concrete example the release names: working with New York’s new DIGIT office to review the safety frameworks AI developers publish and recommend where those frameworks should be strengthened.
The center’s reach has a real limit. It can only evaluate what AI companies choose to disclose publicly, since New York has no authority to compel access to a company’s internal model weights or training data.
What the RAISE Act now requires of AI companies
NYCAIRR sits alongside New York’s RAISE Act, which Hochul signed in 2025 and which DFS is implementing through a newly created office called DIGIT (Office of Digital Innovation, Governance, Integrity, and Trust), the first office of its kind in the country. DFS Acting Superintendent Kaitlin Asrow said DFS is “proud to lead the implementation of the RAISE Act through the new DIGIT Office.”
| Requirement | Deadline | Who it applies to |
|---|---|---|
| Register with New York via the new RAISE Portal | January 1, 2027 | Large frontier AI developers |
| Publish a safety framework on the company’s own website | Starting January 2027 | Large frontier AI developers |
| Report critical safety incidents to the state | Within 72 hours of the incident | All frontier AI developers operating in New York |
| Publish model transparency reports | Ongoing, from registration | All frontier AI developers operating in New York |
The Act does not reach models a company has not yet released, but the disclosures it forces are what NYCAIRR’s researchers will have to work with when they review a developer’s safety claims.
Why New York is doing this instead of Washington
SUNY Chancellor John B. King Jr. and Binghamton University President Anne D’Alleva both framed the launch as New York stepping into a gap left by federal inaction on AI oversight, echoing Hochul’s own language in the release. The center’s structure, a university-hosted research institute feeding assessments to a financial regulator’s new oversight office, has no direct precedent among US states.
Hochul said she will keep looking for ways to build on the RAISE Act and strengthen the state’s AI safety rules in the coming months, which suggests NYCAIRR’s current mandate is a starting point rather than a finished framework. It follows New York’s AI companion chatbot child safety law from October, part of the same push to regulate AI systems the state says the federal government has left unaddressed.
Frequently asked questions
What is the RAISE Act?
A 2025 New York law requiring large frontier AI developers to register with the state, publish safety frameworks, and report critical safety incidents within 72 hours, enforced through the state’s new DIGIT office.
When do AI companies have to register with New York?
Large frontier AI developers must register through the new RAISE Portal by January 1, 2027, after which they must publish safety frameworks and transparency reports.
Who runs NYCAIRR?
Cornell Tech Dean Greg Morrisett and Columbia University research chief Jeannette Wing are interim co-directors, working with more than 80 scientists from New York institutions, backed by a $30 million gift led by Tom Secunda and a $25 million SUNY investment.
What to watch next: whether DIGIT’s first reviews of AI developers’ safety frameworks, due once registrations open on January 1, 2027, lead to New York requiring changes companies have resisted making anywhere else.
Leave a Reply