New York passes RAISE Act to regulate frontier AI safety

New York has taken a historic step in AI governance by passing the Responsible Artificial Intelligence Safety and Education (RAISE) Act, a bold initiative aimed at regulating “frontier AI” models built by powerful tech companies.

According to TechCrunch, the bill now awaits Governor Kathy Hochul’s decision.

The RAISE Act focuses on large-scale AI models, those trained using over $100 million in computing resources and accessible to New Yorkers.

Its chief goal is to prevent catastrophic scenarios, such as AI-driven disasters causing over 100 deaths or $1 billion in damages.

Key provisions include:

• Mandatory publication of safety and security plans with annual third-party reviews.
• Timely reporting of critical incidents, like malicious use or harmful behaviors, to the New York Attorney General.
• Civil penalties up to $30 million for companies that fall short of standards.

Senator Andrew Gounardes emphasized that the bill aims to “ensure AI can flourish while requiring the largest companies to have a safety plan so their products aren't used to hurt people”.

New Accountability Standards for AI Labs

If signed into law, the RAISE Act would require major AI developers, like OpenAI, Google, Anthropic, DeepSeek, and Alibaba, to embrace transparency through:

• Disclosure of detailed safety protocols.
• Incident tracking, including theft of models or misuse for cybercrime or weaponization.

Advocates, including Nobel laureates Geoffrey Hinton and Yoshua Bengio, hail the bill as a landmark for public safety in AI development.

Meanwhile, Cornell professor Lionel Levine noted with alarm that New Yorkers still require more robust oversight for AI than for street vendors, praising the RAISE Act as a much‑needed “common‑sense” safeguard.

Navigating Criticism and Concerns

Despite widespread support, critics warn of unintended consequences:

• Some industry voices, including Andreessen Horowitz, dismiss it as a “stupid, stupid” state-level hurdle that could drive innovation away.
• Anthropic remains concerned about the bill's wide reach, worrying it may inadvertently hamper smaller AI firms, although supporters counter that the law targets only frontier-scale developers.
• Opinion pieces, like one from Paul Lekas, argue that state-specific regulation may hinder innovation and advocate for federal oversight instead.
  
  

Having passed both legislative chambers, the RAISE Act is now with Governor Hochul. She has the power to sign it into law, request amendments, or veto it. If enacted, New York would become the first U.S. state to legislatively mandate transparency, safety planning, and major incident reporting from frontier AI labs.

A Blueprint for Future Governance

The RAISE Act represents a pioneering move in balancing innovation with responsible oversight. Whether it becomes a model for other states or even the federal government will largely depend on Governor Hochul’s action in the next few weeks.

Nutshell: New York’s RAISE Act aims to navigate the frontier of AI safety, demanding accountability from the biggest tech players while safeguarding innovation. Now, all eyes are on the Governor’s decision.