New York Bill Would Let State Limit Compute for Risky AI Models

New York State Senator Andrew Gounardes and Assemblymember Jordan Wright announced in a press release a bill that would expand the state's RAISE Act. The proposal, introduced in the New York Legislature, would let the state limit computing power used to train new models or automate AI research when officials identify catastrophic risks.

Large AI developers would need to report quarterly on automated AI research and computing use. They would also have 72 hours to report substantial increases in risk or a model's ability to automate AI research. The bill would cover companies that spend more than $1 billion on AI development over three years.

The proposal would require annual compliance audits by independent examiners licensed by the state, plus catastrophic risk assessments every 90 days. It would also let the attorney general ask a court to restrict a model's deployment in New York. Penalties could reach $1 million per day for a first violation and $3 million per day for later violations.