New York state lawmakers have passed the RAISE Act, a bill designed to regulate frontier AI models developed by companies like OpenAI, Google, and Anthropic. The legislation aims to prevent these advanced AI systems from causing catastrophic outcomes, defined as events resulting in over 100 deaths or injuries, or more than $1 billion in damages.
Key Provisions of the RAISE Act
The bill establishes America’s first legally mandated transparency standards for frontier AI labs. If signed into law by Governor Kathy Hochul, it would require major AI developers to publish comprehensive safety and security reports about their models. The legislation also mandates reporting of safety incidents, such as concerning AI behavior or model theft.
Transparency Requirements
- Applies to AI models trained using over $100 million in computing resources
- Requires safety and security reports from major AI labs
- Mandates reporting of safety incidents
Industry Reaction
The bill has received support from AI safety advocates, including Nobel laureate Geoffrey Hinton and researcher Yoshua Bengio. However, Silicon Valley has expressed significant resistance, with critics arguing it could stifle innovation. Andreessen Horowitz general partner Anjney Midha called it a “stupid, stupid state level AI bill.”
Addressing Concerns
Co-sponsor Senator Andrew Gounardes emphasized that the bill was designed to avoid chilling innovation among startups and academic researchers, a criticism leveled at similar legislation like California’s SB 1047. Assemblymember Alex Bores argued that the regulatory burden is relatively light and shouldn’t cause companies to stop operating in New York.
Potential Impact
If enacted, the RAISE Act could empower New York’s attorney general to impose civil penalties of up to $30 million on non-compliant tech companies. The bill’s focus on transparency and safety reporting represents a significant step in the AI safety movement, which has faced challenges in recent years.
The passage of the RAISE Act through the New York legislature marks a crucial development in the ongoing debate about AI regulation. Its fate now rests with Governor Hochul, who will decide whether to sign it into law, request amendments, or veto it entirely.
