New York has launched the nation’s first state-run artificial intelligence testing program, marking a significant departure from the federal government’s reliance on industry self-regulation. Governor Kathy Hochul announced the initiative on Tuesday, establishing a framework where independent researchers will evaluate the safety and reliability of advanced AI systems rather than accepting corporate disclosures at face value.
Binghamton University Leads Independent Verification
The operational hub for this effort is the New York Center for AI Responsibility and Research, located at Binghamton University. More than 80 scientists from the institution will conduct the evaluations. To support the computational demands of testing large language models and other complex systems, the center will utilize processing power from Empire AI, a $500 million public-private partnership established by the state.
According to Hochul’s office, the center’s mandate extends beyond simple safety checks. Researchers will analyze how AI technologies can be applied responsibly in high-stakes sectors such as medicine and scientific discovery. The goal is to generate concrete evidence that informs public policy decisions, ensuring that state officials have access to verified data rather than marketing materials.
“The stakes are too high to simply take AI companies at their word,” Hochul stated in a press release regarding the launch. The program aims to fill the regulatory void left by the absence of comprehensive federal oversight, positioning New York as a pioneer in state-level AI governance.
RAISE Act Mandates Corporate Transparency
The testing program operates in tandem with the RAISE Act, legislation signed by Hochul last year. This law requires large AI developers to register with the state by January 1, 2027. Once registered, companies must publish detailed safety frameworks and transparency reports. They are also required to report critical safety incidents within 72 hours of occurrence.
While the RAISE Act increases the volume of data available to researchers, it does not eliminate all blind spots. The state program can only evaluate AI models that companies choose to disclose publicly. Unreleased or proprietary models that developers keep secret will remain outside the scope of these independent tests. However, the mandatory reporting requirements are expected to force a higher level of accountability from major technology firms operating in the state.
The initiative represents a direct challenge to the current status quo, where AI developers largely dictate the terms of their own safety assessments. By establishing a dedicated state facility for verification, New York intends to create a standardized benchmark for AI reliability that could influence national policy discussions in the coming years.
Source: Engadget
