New York Launches Public AI Test and Tightens Transparency Requirements

New York State launches a public initiative to evaluate AI models with independent researchers. Registration and reporting requirements will take effect on January 1, 2027, while the center will rely on the Empire AI infrastructure.

Mandatory Transparency by 2027 and Limited Scope to Published Models

The RAISE Act requires major AI developers to register with the state by January 1, 2027. By this date, they must publish safety frameworks, transparency reports, and report any critical security incidents within 72 hours. However, the law does not cover unpublished models, although it will compel companies to disclose more information. Additionally, the program is limited to studying what actors make public, which poses a potential weakness.

A Center in Binghamton Responsible for Testing and Evidence for Decision-Makers

The New York Center for AI Responsibility and Research, based at Binghamton University, will lead the evaluations. According to the governor's office, it will test advanced AI systems, analyze their risks, and produce evidence for public officials. The center will also explore responsible uses of AI in medicine and scientific discovery.

Launch by Hochul and Resources Mobilized, Including $500M from Empire AI

Governor Kathy Hochul launched this public program on November 10, 2026, touted as the first of its kind at the state level. More than 80 scientists from New York will contribute and leverage the computing power of Empire AI, a public-private partnership endowed with $500 million. "The stakes are too high to simply take AI companies at their word," Hochul stated. The initiative calls for independent researchers to verify the work of AI companies rather than relying on self-regulation.