Overview
Signed by Governor Kathy Hochul on December 19, 2025, the Responsible AI Safety and Education Act (RAISE Act) establishes New York as the second U.S. state after California to enact comprehensive legislation regulating the safety and governance of frontier AI models. The law takes effect January 1, 2027.
Update (March 2026): A negotiated chapter amendment passed the Legislature on March 11, 2026 and was signed by Governor Hochul on March 27, 2026, finalizing the Act. The final version retains the safety-protocol publication requirement and 72-hour incident reporting, and houses the new AI oversight office within the New York Department of Financial Services. The effective date remains January 1, 2027.
Scope
The RAISE Act applies only to "large frontier developers" — entities with more than $500 million in annual gross revenue that train or initiate the training of frontier models exceeding 10²⁶ integer or floating-point operations. In practice this covers a handful of major developers (OpenAI, Anthropic, Meta Platforms, Google DeepMind, xAI, and similar). It does not regulate downstream deployers or the vast majority of AI applications, including most physical security analytics.
Core Requirements
- Publish written safety and security protocols for frontier model development and deployment.
- Report safety incidents to the New York Department of Financial Services (DFS) within 72 hours of determining an incident occurred.
- Submit to oversight by a new AI oversight office within DFS, which will issue annual public reports on frontier developer compliance and risk posture.
- Maintain documentation of model evaluations, red-teaming, and risk-mitigation measures.
Why It Matters for Physical Security Programs
While the RAISE Act does not directly regulate AI video analytics, weapons detection, or workplace violence prevention systems, its disclosure and incident-reporting framework will shape the broader compliance environment for AI in New York. Hospitals, schools, transit authorities, and commercial real estate operators that procure AI security analytics from frontier-class vendors should expect contractual and vendor-management implications, including:
- Vendor due diligence requirements that reference published safety protocols.
- Faster incident-disclosure timelines that may cascade through enterprise security agreements.
- Increased scrutiny on AI model provenance, training data, and model evaluations in procurement.
Compliance Timeline
- December 19, 2025 — Signed by Governor Hochul.
- January 1, 2027 — Operative date for safety protocol publication and incident reporting.
Enforcement
The New York Department of Financial Services is empowered to investigate non-compliance, levy civil penalties, and refer matters to the Attorney General. The statute creates an annual public reporting cadence that will surface failures to publish safety protocols or report incidents in a timely manner.
Related Tracker Entries
- California Transparency in Frontier Artificial Intelligence Act (SB 53, 2026)
- Colorado AI Act Repeal and Replace (SB 26-189, 2026)
- EU AI Act High-Risk Systems Deadline (Annex III, Revised May 2026)
Bill text and history: NY S6953B (Senate, 2025–2026 session)