NEW YORK, November 11, 2026 – Here’s a number worth remembering: 72. That is the number of hours the biggest AI manufacturers will have to give notice about any critical safety incidents to New York State once the new regulations take effect. But the New York State has revealed who will review those notifications.
The Governor Kathy Hochul officially opened the New York Center for AI Responsibility and Research on November 10 in Albany. The center is hosted by Binghamton University, which is a part of the State University of New York (SUNY). The center will bring together more than 80 scientists from various colleges in the state and test the advanced AI systems for their safety and security.
What the center will actually do
Computing brainpower is supplied by Empire AI, a $500 million public-private AI research partnership led by the state. As defined by the governor’s office, the mandate is composed of three components: independent testing of frontier AI, foundational research, and training of the next generation of AI safety experts. Hochul made clear the importance of the mission, stating “the stakes are too high to just take AI companies at their word.” She further took a shot at the Federal government for failing to act. This is Hochul’s narrative; it should be noted that the Federal government was not involved in the announcement.
There are two big players who will manage the operation on an interim basis. Both Greg Morrisett, dean and vice provost of Cornell Tech, and Jeannette Wing, executive vice president for research at Columbia University, have had long careers researching the workings and failures of complex computing systems. The scientists retain their home-campus posts, thus building a statewide network and not a lab.
Hiring begins immediately. Some areas of expertise the center will be looking for include interpretability, alignment, hardware security, agentic systems, formal verification, biosecurity, and many others.
Follow the money
This didn’t happen out of thin air. The idea for the center was first mentioned in January, supported by a $30 million donation from philanthropists headed by Tom Secunda, an alumnus of Binghamton University, along with a $25 million research grant from SUNY.
The rules behind the research
The latter part of Tuesday’s statement was regulatory in nature. According to the RAISE Act in New York, large frontier AI companies will have one year to register via a portal created by the state up till January 1, 2027. Thereafter, they will need to publicly release their safety frameworks through their websites, while all frontier AI companies will need to report on critical events and release model transparency reports. DIGIT is the name of the state office that will do this overseeing.
So does it work? It’s too early to tell for sure. One tech news source pointed out a potential problem: the center would only be able to analyze those things that the company decides to share publicly. Additionally, some policy experts are skeptical about the extent to which the new bill empowers executive agencies. The struggle between state and federal AI regulations is far from being over. The exact extent of access that the center’s test-takers will have to the models remains unclear from the documents I looked through.
Hochul pledges that she would continue to search for other ways to improve the state’s AI legislation in the coming months.
Currently, the plan is straightforward but not easy to implement. Stop trusting assurances; start trusting tests. The deadline of 72 hours has already been set.
