What happened
New York’s Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT) opened an online registration portal on November 10 for “large frontier AI developers” that meet a $500 million revenue threshold. The portal implements the RAISE Act, signed in December 2025 and amended in March 2026, which requires these companies to file a disclosure statement, pay a pro‑rata fee, and submit quarterly risk summaries. Non‑compliance can trigger daily fines of $1,000 and civil penalties up to $3 million per violation. On the same day, Governor Kathy Hochul announced the New York Center for AI Responsibility and Research (NYCAIRR), a $55 million state‑backed institute housed at SUNY Binghamton. NYCAIRR will advise the state on policy, review developers’ safety frameworks, and coordinate with DIGIT, but is exempt from the RAISE reporting requirements. The center’s interim co‑directors are Cornell Tech Dean Greg Morrisett and Columbia EVP for Research Jeannette Wing, and it received $25 million from the state plus $30 million from private donors led by Bloomberg co‑founder Tom Secunda.
On November 10, the Department of Financial Services’ DIGIT unit launched a web‑based portal where AI firms with at least $500 million in annual revenue must register by January 1, 2027, the date the RAISE Act takes effect. The registration requires a detailed disclosure statement covering the company’s New York offices, three points of contact, and any owners holding 5 percent or more of the company.
The RAISE Act also obliges these developers to publish a safety framework and submit a confidential quarterly summary of internal risks. Failure to file a complete or truthful statement can result in a $1,000 per‑day fine, while the Attorney General may pursue civil penalties of $1 million for a first violation and $3 million for subsequent violations.
Simultaneously, Governor Hochul announced the New York Center for AI Responsibility and Research (NYCAIRR), a $55 million institute housed at SUNY Binghamton. The center, co‑directed by Greg Morrisett and Jeannette Wing, will provide expertise to inform state policy, review developers’ safety frameworks, and coordinate with DIGIT on regulatory recommendations. Funding includes $25 million from the state budget and $30 million from private donors led by Tom Secunda.
The RAISE Act explicitly exempts the Empire AI consortium and its affiliated institute from the reporting requirements, meaning the state‑funded AI network is not subject to the same disclosures it helps shape. This exemption has drawn criticism for potential conflicts of interest.
In addition to the registration requirement, the law mandates that all frontier AI developers—regardless of size—report any critical safety incident within 72 hours. DIGIT can share these reports with other government entities, and the reports are exempt from New York’s public‑records law.
Source details: reclaimthenet.org
Why it matters
The rollout represents the first operational enforcement of a U.S. state‑level AI regulatory regime that targets the most financially powerful AI firms. By mandating disclosure of corporate presence, ownership stakes, and safety practices, New York creates a permanent reporting relationship that could serve as a template for other jurisdictions seeking to monitor and mitigate risks from frontier models. The quarterly risk summaries and 72‑hour critical‑incident reporting impose ongoing accountability, while the hefty civil penalties signal strong enforcement intent. NYCAIRR’s $55 million funding and mandate to shape policy provide the state with in‑house expertise, potentially accelerating the development of technical standards for safety, interpretability, and alignment. However, the center’s exemption from the very rules it helps craft raises questions about regulatory capture and the consistency of oversight. The combined regulatory and research apparatus positions New York as a national leader in , influencing how AI developers design, deploy, and monitor large models across the United States.
The registration and reporting framework creates a legally enforceable mechanism for state oversight of the most powerful AI developers, a step beyond voluntary industry self‑regulation.
By requiring public disclosure of ownership and safety practices, the law aims to increase transparency for policymakers, researchers, and the public, potentially reducing the risk of unchecked deployment of high‑risk models.
The substantial civil penalties and daily fines demonstrate a willingness to impose real financial consequences for non‑compliance, which could deter evasive behavior and encourage proactive safety measures.
NYCAIRR’s establishment provides New York with dedicated expertise to evaluate technical safety concerns, potentially accelerating the creation of standards for model interpretability, alignment, and biosecurity.
The exemption for the state‑funded Empire AI consortium raises concerns about regulatory capture, which could undermine the credibility of the oversight regime if the exempt entity influences rule‑making without being subject to the same rules.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
Why can ethical evaluation not be reduced to one model score?
What to watch next
Key developments to monitor include: (1) the first batch of registration filings and any early compliance challenges; (2) how DIGIT enforces the daily fine and civil penalty provisions, especially against high‑profile firms; (3) NYCAIRR’s initial safety‑framework reviews and any recommended rule changes; (4) potential legal challenges from AI companies alleging overreach; and (5) whether other states adopt similar registration and reporting mechanisms, creating a de‑facto national standard.
Early registration filings: Which companies register first, and whether any major players seek extensions or challenge the revenue threshold.
Enforcement actions: Whether DIGIT issues fines or civil penalties in the initial months, and how quickly they respond to missing or inaccurate disclosures.
NYCAIRR output: The first set of safety‑framework reviews, any public recommendations, and whether those lead to amendments of the RAISE Act or new regulations.
Legal challenges: Potential lawsuits from AI firms alleging that the registration and reporting requirements violate interstate commerce or other legal principles.
Policy diffusion: Whether other states or the federal government adopt similar registration portals or reporting mandates, creating a broader regulatory landscape.