New York lawmakers want the power to tell the biggest AI labs to slow down. State Senator Andrew Gounardes and Assemblymember Jordan Wright have introduced a “RAISE Act 2.0” that builds on the AI safety law New York passed last year.
What’s confirmed. The bill is S.10731. Its legislative record shows it was referred to the Senate Rules Committee on October 28, with Senator Patricia Fahy as a co-sponsor. It is a proposal, not a law.
What the bill would do, per the senator’s release:
- Let the state’s new AI office, DIGIT, limit the computing power large developers use to train new models or automate AI research, to reduce catastrophic risk.
- Let the Attorney General ask a court to halt dangerous activity, including restricting a model’s deployment in New York.
- Cover any company spending more than $1 billion on AI development over three years.
- Require quarterly reports on how much AI research is automated, plus 72-hour alerts for big jumps in risk.
- Require each large developer to hire a state-licensed outside examiner.
- Raise penalties to as much as $1 million a day for a first violation.
The framing. Sponsors point to recent incidents where AI models hacked systems on their own. Privacy Daily quotes Fahy saying the bill includes a “kill switch” for models that pose catastrophic risk. That is the sponsors’ case, not a neutral finding.
Our take. This goes further than New York’s current law or California’s SB 53, which mostly rely on disclosure. Machine.net notes that California’s SB 1047, which also reached into how labs run their systems, was vetoed in 2024. Expect a big industry fight.
What’s real here: a bold proposal that is only in committee, a long way from law.
Sources
- Senator Gounardes community portal: RAISE Act 2.0 release
- Machine.net: New York’s Next AI Bill Would Let Regulators Order Labs to Throttle Their Own Training Runs
- Billy: S 10731 legislative record
- Privacy Daily: Reacting to AI Incidents, New York Senator Tees Up Raise Act 2.0
- NYSenate.gov: Sen. Gounardes Proposes Major New AI Safety and Transparency Regulations
By Willa Ravenwire · The AI Bandwagon