New York’s Proposed AI Regulation: The Responsible AI Safety and Education (RAISE) Act
New York is poised to introduce one of the first comprehensive laws regulating advanced AI models in the United States with the Responsible AI Safety and Education (RAISE) Act. This significant legislation has already passed the state Senate and awaits the final decision from Governor Kathy Hochul.
Key Provisions of the RAISE Act
If enacted, the RAISE Act will impose strict safety measures on major AI developers, including OpenAI, Google, and Anthropic. The law mandates that these companies publish detailed safety protocols and conduct risk assessments before releasing advanced AI models to the public. Additionally, developers will be required to report serious incidents, such as model theft or instances of dangerous AI behavior. Violations of these regulations could result in civil penalties reaching up to $30 million.
Targeting Large Corporations
Senator Andrew Gounardes, one of the bill’s sponsors, emphasized that the law is specifically designed to target large corporations that dedicate more than $100 million to model training. Notably, startups and academic institutions will be exempt from these regulations. Gounardes stated, “The window to put guardrails in place is closing fast,” underscoring the urgency of the legislation.
Avoiding Previous Pitfalls
The RAISE Act was crafted to learn from previous legislative efforts, particularly California’s SB 1047 bill, which faced criticism for being overly restrictive. Unlike its predecessor, the RAISE Act does not mandate the implementation of kill switches or hold companies liable for models that undergo modifications.
Industry Response and Concerns
The proposal has drawn criticism from figures in the tech industry, with Anjney Midha, a partner at Andreessen Horowitz, labeling the bill as “dumb.” He cautioned that it could place the U.S. at a competitive disadvantage in the global tech landscape. Jack Clark, co-founder of Anthropic, also expressed concerns about the potential impact on smaller companies, indicating that the legislation might unintentionally stifle innovation.
Despite the pushback, Gounardes defended the bill, asserting that it is intentionally focused on the largest players in the market. Major tech firms such as OpenAI, Google, and Meta have not publicly commented on the bill, but some critics speculate that companies may opt not to offer their models in New York, which could significantly impact the state’s tech ecosystem. Gounardes believes this scenario is unlikely, as New York represents the third-largest economy in the U.S., making it economically unfeasible for companies to withdraw.
Potential Federal Legislation
In a related development, the U.S. House of Representatives recently passed a legislative package that could impose a ten-year ban on state-level AI regulations. If this “One Big, Beautiful” bill passes the Senate, it could effectively block laws like the RAISE Act in the future. Supporters of the moratorium, which includes major tech companies and free-market think tanks, argue that unified federal regulations are necessary to streamline the development of AI technologies. However, critics warn that such centralization could undermine consumer protections and favor industry interests over public safety.
Conclusion
The RAISE Act represents a significant step towards establishing regulatory frameworks for advanced AI technologies in New York. By mandating safety protocols and risk assessments, the legislation aims to balance innovation with public safety in an increasingly complex technological landscape. As the bill awaits the governor’s signature, its implications will likely resonate across the tech industry and set a precedent for AI regulation nationwide.
