NY Passes AI Safety Bill to Prevent Disasters
New York Takes Lead on AI Regulation
New York lawmakers have approved the RAISE Act, groundbreaking legislation aimed at preventing catastrophic scenarios from frontier AI models developed by tech giants like OpenAI, Google, and Anthropic. The bill specifically targets potential disasters involving 100+ casualties or $1+ billion in damages.
Key Provisions of the RAISE Act
- Transparency requirements: Major AI labs must publish detailed safety reports
- Incident reporting: Mandatory disclosure of safety breaches and model theft
- Financial penalties: Up to $30 million fines for non-compliance
Balancing Innovation and Safety
Unlike California’s vetoed SB 1047, the RAISE Act was carefully designed to avoid stifling innovation. Senator Andrew Gounardes emphasized that smaller companies and academic researchers wouldn’t face undue burdens.
“The window for AI guardrails is closing fast,” Gounardes warned. “AI experts confirm these risks are real and alarming.”
Industry Pushback and Responses
The bill has faced criticism from Silicon Valley investors like Andreessen Horowitz, who called it “stupid” and harmful to U.S. competitiveness. Even safety-focused Anthropic expressed concerns about potential impacts on smaller firms.
Lawmakers counter these arguments by noting:
- The bill only applies to models trained with $100+ million in computing resources
- New York’s massive economy makes withdrawal impractical for most companies
- Regulatory burdens are significantly lighter than in European markets
What Happens Next?
The RAISE Act now awaits Governor Kathy Hochul’s decision. If signed into law, New York would establish America’s first legally-mandated AI transparency standards – potentially setting a precedent for federal regulation.
As AI capabilities advance rapidly, this legislation represents a critical test for balancing technological progress with public safety. The tech industry and policymakers worldwide will be closely watching New York’s next moves.