New York RAISE Act sets fines up to $3 million for AI developers
The New York RAISE Act, signed by Governor Kathy Hochul on December 19, 2025, requires large frontier AI developers to publish their safety protocols and report safety incidents to the state within 72 hours. Fines reach $1 million for a first violation and $3 million for later ones. For companies building advanced AI models, it turns safety documentation and incident reporting into legal duties in one of the country's leading tech states.
Why New York passed its own AI safety law
New York acted because, in the governor's words, the federal government lags behind on AI rules. Governor Kathy Hochul said the law builds on California's recently adopted framework and creates a unified benchmark among the country's leading tech states while Washington has failed to put common-sense regulations in place.
The law is the Responsible AI Safety and Education Act, bill S6953B/A6453B. State lawmakers passed it in June 2025. After lobbying from the tech industry, Hochul proposed changes to scale it back; TechCrunch, citing The New York Times, reported that she agreed to sign the original bill while lawmakers agreed to make her requested changes the following year.
"By enacting the RAISE Act, New York is once again leading the nation in setting a strong and sensible standard for frontier AI safety, holding the biggest developers accountable for their safety and transparency protocols." — Governor Kathy Hochul, on signing the law
What the RAISE Act requires from frontier AI developers
Hochul signed the law on December 19, 2025, making New York the second US state after California to enact major legislation on frontier AI safety. It applies to large developers of frontier AI models.
Those developers must create and publish information about their safety protocols. They must also report safety incidents to the state within 72 hours of determining that an incident occurred.
How oversight and enforcement will work
The law creates an oversight office within the New York State Department of Financial Services. The office will assess large frontier developers, promote transparency and issue reports every year.
Enforcement runs through the attorney general, who can bring civil actions against large frontier developers that fail to submit required reports or make false statements. The law does not give private individuals a right to sue, according to the law firm MoFo.
Kaitlin Asrow, acting superintendent of the Department of Financial Services, said the department has been developing rules that help financial services companies adopt AI responsibly.
What penalties companies face under the law
Penalties are up to $1 million for a first violation and up to $3 million for each later violation. The law does not take effect immediately.
The sponsors called it the strongest law of its kind. Assemblymember Alex Bores said it goes beyond California's SB 53 in significant ways. Both OpenAI and Anthropic expressed support for the bill while calling for federal legislation. Opposition remains: a super PAC backed by Andreessen Horowitz and OpenAI president Greg Brockman is looking to challenge Bores. The law also arrived just after President Donald Trump signed an executive order that directs federal agencies to challenge state AI laws.
What the New York RAISE Act means for companies building AI
For a company that develops advanced AI models, the law turns safety documentation and incident handling into legal duties with fines attached. MoFo advises developers to assess whether their models and compute practices could bring them under the law, review their existing safety and governance documents, and check whether they can identify and report a qualifying incident within 72 hours.
The duties fall on large developers. Companies that buy AI models gain something too: the safety protocols those developers must publish, which they can read before choosing a model. Developers also have to watch how the New York rules are applied alongside California's SB 53 and federal activity.
Questions and answers
What are the fines under the New York RAISE Act?
The attorney general can seek penalties of up to $1 million for a first violation and up to $3 million for later violations, for failing to submit required reports or for making false statements.
How fast must AI developers report a safety incident?
Large frontier developers must report an incident to the state within 72 hours of determining that it occurred.
Who oversees the RAISE Act?
A new office within the New York State Department of Financial Services will assess large frontier developers and issue reports every year.
MOZGII view
How we would build it
A law that requires a report within a fixed window is really a requirement to know, quickly, that something happened. For a developer of a large model, the hard part is not writing the report. It is noticing the incident, deciding whether it qualifies and getting the facts to the right people in time. If we were building that system, we would start with an incident register: one record per event, with when it was found, who found it, what model and version it involved and what has been done.
We would route every flag from safety testing, user reports and internal monitoring into that register. A small team would decide within hours whether each event qualifies, and a lawyer would review every report before it goes to the state. The clock would start in the register, so nobody has to reconstruct dates afterward.
It would break on unclear cases. Events that might qualify sit in inboxes while people argue, and the window runs out. It would also break if the published safety protocol and the actual practice drift apart, because a false statement carries its own penalty.
The one number we would watch is the time from first detection to a decision on whether an event must be reported. Next, we would build a single map of each model's duties under the New York and California laws, so one incident does not need two separate processes.
Sources
Where the facts in this story come from.
- 01New York governor Kathy Hochul signs RAISE Act to regulate AI safety | TechCrunchtechcrunch.com
- 02Governor Hochul Signs Nation-Leading Legislation to Require AI Frameworks for AI Frontier Models | Governor Kathy Hochul | New York Stategovernor.ny.gov
- 03New York Enacts the RAISE Act Regulating Frontier AI Modelsmofo.com