New York's ASSET Unit Puts Frontier AI Vendors Under AG Enforcement Before 2027
Source
Attorney General James Announces Creation of AI Safety, Standards, Enforcement, and Transparency UnitNew York Office of the Attorney General
What happened
New York Attorney General Letitia James announced the creation of the AI Safety, Standards, Enforcement, and Transparency (ASSET) Unit within her office. The unit is tasked with enforcing New York's Responsible AI Safety and Education Act (RAISE Act) for Large Frontier Developers and other New York AI-related laws. In September 2026, the AG issued an industry alert inviting workers to submit confidential complaints through a whistleblower portal. The alert targeted workers with knowledge of unsafe or potentially illegal AI development. Under the RAISE Act, which takes effect January 1, 2027, frontier developers must publish transparency reports and disclose Critical Safety Incidents. Large frontier developers face additional obligations, including a published frontier AI framework and disclosures to the New York Department of Financial Services (NYDFS). A new NYDFS office called DIGIT (Office of Digital Innovation, Governance, Integrity and Trust) receives incident reports and risk-assessment summaries. The law directs NYDFS to consider transmitting those materials to the AG. A single incident filing could therefore trigger a parallel ASSET Unit investigation. Penalties for large frontier developers reach $1 million for a first violation and $3 million for each subsequent violation. NYDFS may impose $1,000 per day for disclosure failures. The Act creates no private right of action. The AG's office may also invoke New York's data security and breach-notification law (the SHIELD Act), chatbot safety disclosure requirements, and its general authority over unfair and deceptive practices.
Why it matters
- ·Frontier AI developers with New York operations face a dual-regulator enforcement structure. NYDFS DIGIT receives incident reports and risk assessments. The ASSET Unit can independently investigate based on those same filings. This doubles compliance exposure from a single unreported or misfiled Critical Safety Incident.
- ·The active whistleblower program launched before the New York's Responsible AI Safety and Education Act (RAISE Act) for Large Frontier Developers takes effect. Enforcement intelligence is being collected now. Enterprises using frontier AI vendors should verify those vendors have accurate, defensible transparency reports and anti-retaliation policies in place well before January 1, 2027.
- ·Federal preemption of state AI development laws remains an open question under the current administration's AI policy framework, which calls for preempting state laws that regulate AI development. That uncertainty does not reduce immediate compliance obligations but does create legal exposure planning risk for multi-jurisdiction programs.
Stay ahead of stories like this
Get every US AI governance development like this one, plus the rest of the week's developments. Every Thursday.
Recent issues
- Regulators, courts, and markets are all testing whether AI companies' accountability structures hold up under real pressure, not just in policy documents.19 Nov
- The next quarter will test whether compliance programs can keep pace with AI systems that breach containment, reshape capital markets, and enable fraud at a scale no prior control framework anticipated.12 Nov
Free every Thursday. Unsubscribe anytime.
Governance controls affected
What to do now
- ☐Identify every AI vendor in your stack that could qualify as a frontier developer or large frontier developer under the RAISE Act, and confirm whether each has a published frontier AI framework and Critical Safety Incident reporting process ready for January 1, 2027.
- ☐Ask frontier AI vendors to provide their RAISE Act compliance documentation, including their transparency report and their NYDFS DIGIT disclosure plan, before the effective date.
- ☐Review your own internal AI incident escalation thresholds to confirm that any event qualifying as a Critical Safety Incident under the RAISE Act would be caught, logged, and reported by your vendor within required timeframes.
- ☐Check whether your AI vendor contracts require the vendor to notify you if they receive a regulatory inquiry or whistleblower complaint from the ASSET Unit or NYDFS DIGIT, and add that requirement if it is absent.
- ☐Brief your legal and compliance leads on the federal preemption risk so that your RAISE Act compliance program accounts for the possibility of a legal challenge affecting enforcement timelines, without pausing preparation.
What to watch next
Compliance teams should monitor NYDFS DIGIT rulemaking, which will define Critical Safety Incident reporting procedures and may designate equivalent federal standards that could affect how vendors comply across jurisdictions. After January 1, 2027, the ASSET Unit's first enforcement actions will show how aggressively the AG applies the RAISE Act's penalty structure. They will also reveal whether whistleblower tips translate into formal investigations. Separately, federal action to preempt state AI development laws could affect the RAISE Act's enforceability. Teams should track that litigation risk alongside their compliance preparation.
