New York Opens Registration for Frontier AI Labs and Launches a Safety Research Center

New York has opened a registration portal for large frontier AI developers, including Anthropic and OpenAI, and Governor Kathy Hochul has launched the New York Center for AI Responsibility and Research at SUNY Binghamton University. The Center's stated purpose is to advance AI science and safety (Governor's office). Both steps are part of the state's implementation of the RAISE Act, as TechCrunch reports.

Large frontier developers, meaning companies building the most capable general-purpose AI models, must register with the portal by January 1, 2027. The RAISE Act defines a large frontier developer as a company with more than $500 million in annual revenue. The state's Office of Digital Innovation, Governance, Integrity, and Trust, known as DIGIT, will implement the law and oversee these labs.

The act's obligations are specific. Covered developers must publish safety frameworks on their websites, issue public model transparency reports, and report critical safety incidents within 72 hours. Penalties can reach $1 million for a first violation and $3 million for subsequent violations.

The Center is designed to sit alongside DIGIT rather than replace it. It will independently evaluate the safety, security, and operation of advanced AI systems, and it will work with DIGIT to review companies' safety frameworks. Hochul first announced the Center in January, with $55 million in donations and capital behind it.

New York is also enforcing a separate law covering AI companions, the chatbots that hold conversations with users. Operators must disclose to users that they are bots and make reasonable efforts to detect suicidal ideation and self-harm. A companion chatbot safety law for minors takes effect June 1, 2027.

Other states are moving on parallel tracks. California's SB 53 requires frontier AI labs to publish safety frameworks, enact whistleblower protections, and report critical safety incidents within 15 days. Illinois's SB 315 takes effect at the start of 2027 and includes independent safety compliance audits.

The clearest point of divergence is the incident reporting window: 72 hours under New York's RAISE Act, 15 days under California's SB 53. A lab operating in both jurisdictions will have to route one incident through two clocks. How hard that is in practice will depend on how each state defines a critical incident.

The broader context here is that frontier labs now face overlapping state obligations covering frameworks, transparency, incident reporting and, in Illinois's case, audits. For engineering and compliance teams, the practical question is less whether these rules apply than whether one set of internal evaluation, logging and escalation processes can satisfy each regime's definitions and deadlines without duplicating work.

In my view, the Center's independent evaluation mandate is the element most likely to shape how these rules work in practice. A public, university-based body that can examine model safety and security directly, and review company frameworks alongside a state regulator, gives DIGIT technical capacity it would otherwise have to build or contract out. Whether that capacity is used well will depend on access, staffing and the depth of the evaluations, none of which the announcement settles.

The January 1, 2027 registration deadline, and the chatbot dates that follow it, will provide the first measurable signal. Developers will need to show what they have published and how they have classified incidents, and regulators will need to show they can read it.