New York is the second U.S. state to pass leading AI safety legislation, behind only Washington state, under a new law called RAISE, which Governor Kathy Hochul signed.
The state Legislature adopted the RAISE Act in June, but after tech industry lobbying, Kathy Hochul made proposals to dilute the bill. Kathy Hochul eventually agreed to sign the original bill, but lawmakers also agreed to make the changes she wanted next year, The New York Times reports.
The bill would require large AI developers to disclose the safety of their products and report safety incidents to the state within 72 hours. It will also establish a new office within the Department of Financial Services to oversee the development of AI.
If companies fail to submit safety reports or lie, they can face fines of up to $1 million ($3 million for repeat offenses). California Governor Gavin Newsom signed a similar safety bill in September, a date Kathy Hochul cited in her announcement.
“This law joins a recently enacted framework in California to establish a common benchmark among the nation’s principal tech states as the federal government has been slow to act on simple regulations that keep us safe,” Kathy Hochul said.
State Senator Andrew Gounardes, one of the sponsors of the bill, wrote: “#BigTech thought they could weasel their way into killing our bill. We blocked them and enacted the strongest AI safety law in the nation.”
OpenAI and Anthropic both told the NYT they supported New York’s bill. Still, they said there was a need for federal legislation, with Sarah Heck, Anthropic’s head of external affairs, saying: “Two of the largest states in our country have now enacted AI transparency bills, which indicates the critical importance of safety and should encourage Congress to build upon them.
Not everybody in the tech industry has been so soft. Indeed, a super PAC funded by Andreessen Horowitz and OpenAI President Greg Brockman is aiming to challenge the man who co-sponsored that bill with Gounardes, Assemblyman Alex Bores. (“I think the fact that they’re being candid about it’s good,” Bores told reporters.)
That would come after President Donald Trump signed an executive order directing federal agencies to supersede state AI laws. The order — which is supported by Trump’s AI czar, David Sacks — is the Trump Administration’s latest attempt to preempt states from regulating AI and will almost certainly be challenged in court.
We also discussed Trump’s executive order, as well as the anti-state-regulation-of-AI work Sacks and a16z have been engaged in, on this week’s episode of the Equity podcast.
