November 11, 2026 – New York has opened an online registration portal for developers of advanced artificial intelligence (AI) models, giving major technology companies a way to prepare for new safety and transparency rules. Governor Kathy Hochul’s office confirmed the portal’s launch on November 10. Large AI developers covered by the state’s RAISE Act must register by January 1, 2027.
The portal is part of New York’s Responsible AI Safety and Education (RAISE) Act, which was signed into law in December 2025. It is overseen by the Office of Digital Innovation, Governance, Integrity and Trust (DIGIT), part of the New York State Department of Financial Services. According to the state’s official AI oversight information, the main requirements take effect on January 1, 2027. Opening the portal does not mean those requirements are already being enforced.
The law does not apply to every AI chatbot or technology startup. It focuses on powerful AI systems called frontier models, which meet a computing threshold of more than 10²⁶ operations during training and related development. A developer is considered large when it meets that technical threshold and its business, together with affiliated companies, earned more than $500 million in gross revenue during the previous calendar year.
The new registration process is relevant to major AI developers such as OpenAI and Anthropic, although whether a company must register depends on the law’s specific requirements. Under the official registration guidance, an authorized company representative must use an NY.gov account, establish permission to act for the developer and provide business information. Required details include company addresses, ownership information where applicable, and three points of contact. Registrations generally need renewal every two years or sooner when certain important details change.
The RAISE Act also establishes reporting responsibilities that extend beyond large developers. All covered frontier AI developers operating in New York must publish information about new or substantially modified models and report qualifying critical safety incidents. Such incidents must be reported within 72 hours after the developer determines an incident occurred or has sufficient reason to believe one occurred. An incident posing an immediate risk of death or serious physical injury also requires notification to an appropriate authority within 24 hours.
Large developers face additional responsibilities. They must publish AI safety frameworks explaining how they identify serious risks, protect unreleased models, respond to dangerous incidents and oversee internal safety procedures. They must also submit summaries of their assessments of potentially catastrophic risks every three months. These requirements are intended to make the development of powerful AI systems more transparent and accountable.
The law includes financial consequences for certain violations. When Governor Hochul signed the RAISE Act, her office confirmed that the state could seek civil penalties of up to $1 million for a first violation and up to $3 million for subsequent violations involving failures to submit required reports or false statements. These are possible enforcement penalties, not charges automatically imposed when a company registers.
Alongside the portal, New York launched the Center for AI Responsibility and Research at Binghamton University, an initiative first announced in January 2026. The center aims to bring together more than 80 researchers from institutions across the state to independently evaluate advanced AI systems, study security risks and develop better testing methods. Its research could help DIGIT assess developers’ published safety frameworks and strengthen future oversight.
For AI developers, security teams and businesses that rely on advanced models, the change creates practical responsibilities. Companies will need to determine whether their systems meet New York’s legal definitions, establish reliable incident-reporting procedures and prepare accurate safety documentation. Enterprise customers may also benefit from greater access to information about how AI developers identify and manage risks. However, registration alone should not be treated as proof that a model is safe.
New York’s move comes as governments consider different approaches to AI oversight. TechBonafide recently reported on Australia’s proposed AI safety regulator, which is not yet operating under the proposed legislation. New York’s position is different because its RAISE Act has already been enacted, although the main requirements are scheduled to begin in January 2027.
Some important questions remain unanswered. Publicly available official information does not confirm whether OpenAI, Anthropic or other named developers have completed registration. It is also too early to know how effective the reporting and oversight system will be once the requirements take effect. For now, the confirmed development is that New York has opened the registration process and established a clear compliance deadline for covered large AI developers.
