New York Opens RAISE Act Registration for Large Frontier AI Developers

New York Governor Kathy Hochul announced the registration opening on November 10, 2026 for large frontier artificial intelligence developers covered by the Responsible AI Safety and Education Act. Filings are due by January 1, 2027, when the law takes effect. The registration requirement applies to developers that, together with their affiliates, had more than $500 million in annual gross revenue in the preceding calendar year.

The Act covers frontier models developed, deployed or operating wholly or partly in New York. A frontier model crosses the statutory threshold when trained using more than 10^26 integer or floating-point operations. Oversight sits with the new Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT), within the Department of Financial Services.

Registration and Safety Duties

A large frontier developer must maintain a current disclosure statement and pay its required assessment to develop, deploy or operate a covered model in New York. Renewal is required every two years, upon a frontier-model ownership transfer or upon a material change to reported information, whichever comes first.

Filings must identify business names and addresses, three regulatory contacts and specified beneficial owners. DIGIT will publish a list of registrants without their regulatory contact information, while the Department of Financial Services will assess developers pro rata to cover the direct and indirect costs of administering the Act.

Large developers must also write, implement, follow and publish a frontier AI framework addressing catastrophic-risk thresholds, mitigations, deployment reviews, third-party assessments, security for unreleased model weights, incident response and internal-use risks. Frameworks require at least annual review; material modifications and their justification must be published within 30 days.

The safety duties are built around catastrophic risk, defined as a foreseeable and material risk of more than 50 deaths or serious injuries, or more than $1 billion in property damage or loss, from a single incident involving specified weapons assistance, qualifying autonomous conduct or evasion of developer or user control.

Before or alongside deployment of a new or substantially modified model, every frontier developer must publish a transparency report covering its release date, supported languages, output modalities, intended uses and generally applicable restrictions. Large developers must add summaries of catastrophic-risk assessments, their results and third-party evaluator involvement. Redactions for trade secrets, cybersecurity, public safety, U.S. national security or legal compliance must be described and justified, with unredacted information retained for five years.

The reporting provisions require a critical safety incident to be reported to DIGIT within 72 hours after the developer determines that one occurred or learns facts sufficient to establish a reasonable belief that one occurred. An incident posing an imminent risk of death or serious physical injury must instead be disclosed within 24 hours to an appropriate law-enforcement or public-safety authority, as required by law. Large developers must also confidentially submit summaries of catastrophic-risk assessments from internal model use every three months, unless DIGIT agrees to another schedule.

The attorney general may seek civil penalties of up to $1 million for a first specified violation and up to $3 million per subsequent violation. Covered conduct includes failures to provide compliant documents or report incidents, prohibited false or misleading statements, and failure to follow a developer’s own framework. Separately, after notice and a hearing, the office may impose $1,000 for each day an entity fails to file a required registration disclosure or correct false information, plus assessments owed.

Research Center Will Support Framework Reviews

Hochul also detailed the implementation role of the New York Center for AI Responsibility and Research at Binghamton University. Greg Morrisett of Cornell Tech and Jeannette Wing of Columbia University will serve as interim co-directors. The state plans for the center to convene more than 80 researchers across New York and use the computing resources of Empire AI, a $500 million public-private research initiative.

Its remit includes independent testing of frontier systems, foundational safety research, development of evaluation measurements and standards, and training AI safety experts. At the state’s request, the center will coordinate with DIGIT to review developers’ published safety frameworks and recommend stronger requirements.

Hochul first announced the center on January 21, 2026, backed by a $30 million philanthropic gift from donors led by Tom Secunda and a $25 million SUNY research capital investment. Accredited New York colleges and universities are exempt from the Act when conducting academic AI research, as are the Empire AI consortium and institute.

Registration is the immediate task, but the law’s lasting operating burden is continuing disclosure: current filings, public safety frameworks and model reports, rapid incident notifications and periodic internal-use assessment summaries. The center’s reviews give DIGIT an independent source of technical scrutiny as that reporting system begins.