New York's AI Safety Law Takes Effect, Putting Banks' Vendor Plans to the Test

New York Governor Kathy Hochul announced the implementation timeline for the RAISE Act on September 21, 2026: starting in November, large frontier model developers must register in the state; from January next year, requirements for public safety protocols, periodic reporting, and major incident notifications will officially take effect. Hochul also appointed Marc Gilman as deputy director of the state's Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT), which operates under the Department of Financial Services.

Under the plan, covered developers must publish a safety and transparency framework, submit catastrophic risk assessments quarterly, and report major safety incidents to DIGIT within 72 hours, while completing registration and periodic disclosures.

These obligations directly bind large model developers and do not automatically apply to banks and fintech companies that procure their services. However, if a bank's customer service, anti-fraud, or payment processes are deeply tied to a vendor's model, and that vendor experiences a safety incident leading to investigations, access restrictions, or changes in model behavior, the bank must still continue serving account holders and processing transactions. Its vendor plans will face a real test.

Source: https://www.pymnts.com/news/artificial-intelligence/2026/new-yorks-ai-safety-law-puts-banks-vendor-plans-to-the-test/