New York’s RAISE Act Would Require Security Protocols from Leading AI Developers

New York state legislators passed the RAISE Act last week, aiming to prevent artificial intelligence (AI) models from contributing to disaster scenarios. It currently awaits a signature from New York Governor Kathy Hochul.

The Responsible Artificial Intelligence Safety and Education (RAISE) Act (S6953B/A6453B) requires the largest AI developers—including Google, OpenAI, and Anthropic—to develop a safety plan to protect against automated crime, bioweapons, and other widespread harm and destruction that could cause death or injury to 100 or more people or more than $1 billion in damages.

The RAISE Act sets three primary safeguards:

• It requires the largest AI companies to publish safety and security protocols and risk evaluations. The protocols must cover severe risks, including the AI systems’ use to create biological weapons or carry out criminal activity. Safety plans would need to be reviewed by a qualified third party.
  
  
• It requires companies to disclose serious incidents, such as if a dangerous AI model is stolen by a malicious actor, to the attorney general and the Division of Homeland Security and Emergency Services.
  
  
• It allows the New York State Attorney General to bring penalties against large AI companies that do not meet the act’s standards.
  
  

The provisions in the act mirror some of those in California’s vetoed AI safety bill, SB 1047. But the New York version was designed to not chill innovation among startups or academic researchers. If signed, the law would only apply to AI companies that have spent more than $100 million in computational resources to train advanced AI models. The RAISE Act also does not require AI model developers to include a “kill switch” on their models, and it does not hold companies that train existing frontier AI models accountable for critical harms, TechCrunch reported.

“Many major AI companies have voluntarily committed to create safety plans, but there is currently no legal requirement that they do so,” according to a press release from one of the act’s co-sponsors, New York State Senator Andrew Gounardes. “By writing these protections into law, the RAISE Act ensures no company is incentivized to cut corners or put profits over safety.”

Silicon Valley influencers, technologists, and AI professionals pushed back on the act, similar to their reaction to California’s SB 1047. Critics also warned that AI model developers could skip offering their most advanced AI models in New York to avoid the regulation.