TL;DR
What did New York announce about AI safety and the RAISE Act?
New York opened a state center to check powerful AI models and a portal where big AI companies must register by January 1, 2027. The AI tools you use will face outside checks and must report serious safety problems. You don't need to do anything yet.
What did Gov. Hochul announce?
On Tuesday, November 10, 2026, Gov. Kathy Hochul launched the New York Center for AI Responsibility and Research, based at SUNY's Binghamton University. The center will bring together more than 80 researchers from colleges and research institutions across the state to evaluate advanced AI systems for safety and security, and to advise state officials on AI policy.
On the same day, the state's new Office of Digital Innovation, Governance, Integrity, and Trust, known as DIGIT, opened an online portal where the largest AI developers must register under the RAISE Act, New York's AI safety law. "The stakes are too high to simply take AI companies at their word," Hochul said.
What will the New York Center for AI Responsibility and Research do?
The center's job is to work out how safe the most powerful AI models really are, rather than relying on what the companies say about them. According to the Governor's office, it will:
- Check frontier AI systems, meaning the most advanced models from companies like OpenAI, Anthropic and Google, for safety and security.
- Develop ways to measure AI, the scientific tools, measurements and standards needed to evaluate AI as it gets more capable.
- Train new AI safety experts and start a national search for researchers in areas like formal verification (mathematically proving software works as intended), interpretability (understanding why a model gives the answers it does) and biosecurity.
- Advise the state. At New York's request, it can give evidence and assessments to inform policy, starting with reviewing AI developers' published safety frameworks with DIGIT and suggesting how to strengthen the rules.
Computer scientists Greg Morrisett, dean of Cornell Tech, and Jeannette Wing of Columbia University will lead it as interim co-directors. The researchers will keep working at their home institutions, using computing power from Empire AI, the state's $500 million public-private AI research project.
Who has to register under the RAISE Act, and by when?
Large frontier AI developers have to register through the new RAISE Portal by January 1, 2027. The law covers all frontier AI developers operating in New York, with extra duties for the large ones.
| Requirement | Who it applies to | When |
|---|---|---|
| Register on the RAISE Portal | Large frontier AI developers | By Jan 1, 2027 |
| Publish a safety framework on their website | Large frontier AI developers | Starting in January 2027 |
| Report critical safety incidents | All frontier AI developers | Starting in January 2027; within 72 hours of an incident |
| Publish model transparency reports | All frontier AI developers | Starting in January 2027 |
A safety framework is a company's written plan for how it checks its models and limits their risks. A transparency report describes what a model is and how it was checked. These rules fall on the companies that build the models, not on the people and businesses that use them.
How is the money set up?
Hochul first announced the center in January with a $30 million gift from donors led by Bloomberg LP co-founder Tom Secunda, a Binghamton graduate, plus an initial $25 million research investment from SUNY and operating money from the state. Tuesday's news adds the statewide team, the leaders and the start of its work.
How does New York's plan compare with Australia's?
New York moved a day after Australia announced its own plan for mandatory safety rules and a new AI safety regulator for frontier AI companies, including incident reporting. Australia's rules still have to be written into law. New York's RAISE Act is already law, and its registration and reporting duties start in January.
Both push in the same direction: the biggest AI makers report problems to the government, and someone outside the company checks their safety claims.
Our take
This is a sensible move. Right now, most of what the public knows about a model's risks comes from the company that sells it. A team of 80-plus university researchers with serious computing power gives the state its own way to check those claims, and the 72-hour incident rule means problems should surface faster.
The open question is enforcement. The release describes the center's role as evidence and recommendations, while DIGIT oversees the companies. How much this changes depends on how DIGIT enforces the RAISE Act from January. For everyday users the effect will be slow and mostly invisible, but over time it should mean earlier public warnings when an AI product misbehaves.
— Artificially Nerdy
What to do this week
Nothing for most readers. The RAISE Act's duties fall on the companies that build frontier AI models, not on people or small businesses that use AI tools. Two things worth keeping an eye on:
- After January, look for the safety frameworks and transparency reports the big AI makers will publish, if you want to know how the tools you rely on were checked.
- Building on top of a big AI model? Follow your provider's safety and incident notices, since New York's rule means serious incidents get reported within 72 hours.
FAQ
Do I have to register my business's AI tools with New York?
No. Registration is for large frontier AI developers, the companies that build the most advanced models. Using ChatGPT, Claude or Gemini at work doesn't put you under the RAISE Act's registration rule.
Where is the new AI center based?
At SUNY's Binghamton University. Its researchers will stay at their own colleges and institutions across New York and work together through the center.
Who runs the RAISE Act day to day?
New York's Department of Financial Services leads it through the new DIGIT office, according to the Governor's release.
Sources
- Office of Governor Kathy Hochul: Governor Hochul Declares NY Will Lead the Nation in AI Science and Safety; Launches New Research Institute — November 10, 2026. Center details, 80+ researchers, co-directors, Empire AI ($500 million), funding ($30 million gift, $25 million SUNY), DIGIT portal, Jan 1, 2027 registration, safety frameworks, incident reports within 72 hours, transparency reports, Hochul quote.
