A woman in a navy jacket bearing a state seal speaks at a lectern with a sign reading Keeping New Yorkers Safe, flanked by three officials, in front of a wall of video monitors showing subway stations.
Governor Kathy Hochul at a press conference in the MTA headquarters situation room with transit and police officials, October 9, 2025. Archive photograph by Marc A. Hermann / Metropolitan Transportation Authority (CC BY 4.0); Hochul's office announced the new AI research center and the RAISE Portal on November 10, 2026. Marc A. Hermann / Metropolitan Transportation Authority / Wikimedia Commons · CC BY 4.0

What New York Announced

On November 10, 2026, Governor Kathy Hochul launched the New York Center for AI Responsibility and Research (NYCAIRR) at SUNY Binghamton University. The same release says the Office of Digital Innovation, Governance, Integrity and Trust (DIGIT), a new office within the Department of Financial Services (DFS), opened the RAISE Portal, where large frontier AI developers operating in New York must register by January 1, 2027 under the state's Responsible AI Safety and Education (RAISE) Act. It adds that Greg Morrisett of Cornell Tech and Jeannette Wing of Columbia University will lead the center as interim co-directors, and that the center will bring together more than 80 scientists across New York State.

A January 21, 2026 release had first announced the center, with a $30 million philanthropic gift from donors led by Tom Secunda and a $25 million SUNY investment.

Filings and Reports DFS Describes

DFS guidance defines a frontier model as a foundation model trained with more than 10^26 integer or floating-point operations, counting later fine-tuning, reinforcement learning, and other material modifications, and a large frontier developer as a frontier developer that, together with its affiliates, had annual gross revenues above $500 million in the preceding calendar year. According to DFS, large frontier developers operating in New York must keep a current registration beginning January 1, 2027, publish a frontier AI safety framework, and summarize for DIGIT, within 30 days of each calendar quarter's end, their assessment of catastrophic risk from internal use of their models. Their transparency reports for a new or substantially modified model must also state “the extent to which third-party evaluators were involved.” The duty is to disclose that involvement, and the guidance does not say an evaluator is required.

Starting January 1, 2027, all frontier developers operating in New York must report a critical safety incident to DIGIT within 72 hours of determining that one occurred or learning facts sufficient to establish a reasonable belief that it did. An incident that poses an imminent risk of death or serious physical injury must be disclosed within 24 hours to an appropriate authority, such as law enforcement. DFS's public-reporting page says the Act's four defined incident types do not cover all AI-related errors, malfunctions, or consumer complaints.

Limits and Audit Implications

The release describes the center's testing only in the future tense. It says the center will pursue independent testing of frontier AI systems and will coordinate with DIGIT to begin reviewing published safety frameworks, and it names no system tested, result, start date, or terms for developer access. Everything here about the center comes from the state's own releases.

Large frontier developers must publish safety frameworks on their own websites, and all frontier developers must publish transparency reports there. Registrations, incident reports, and quarterly summaries are developer-supplied filings to DIGIT, and the DFS pages do not say whether any will be public or how DFS would check them.

Our editorial reading: registration, incident, and evaluator-disclosure duties define records to seek after January 1, 2027, while the center's testing remains a plan until it publishes methods or results.