The Conference of State Bank Supervisors released its Artificial Intelligence Supervisory Framework on 16 September 2026. The CSBS State Supervisory Processes Committee and the CSBS NonDepository Supervisory Committee approved the material in August 2026. CSBS describes the release as a discretionary tool for state examiners rather than a binding rule, and states that each state financial regulatory agency will determine the extent to which it is incorporated into that agency's supervisory programme.
The package is made up of named parts. The Core Examiner Guide covers AI inventory and use cases. The Examiner Work Program sets out the examination steps. The Nonbank AI Supplements add review points for third-party and vendor risk at nonbank financial services companies. The AI Use Case Risk Tiering Worksheet is presented as an optional industry tool supporting a risk-based assessment of individual AI use cases. A Source Support Document accompanies the set. There is no statutory instrument behind the release and no section or rule number to cite.
State-chartered banks and state-licensed nonbank financial services companies are the institutions examiners will assess using this material. A firm running AI in any regulated activity should expect an examiner to ask for an inventory of its AI use cases and to work through risk tiering on individual deployments. Technology vendors supplying AI systems to those firms come into scope through the third-party and vendor review points in the Nonbank AI Supplements.
Adoption is not uniform across the states. Because each agency sets its own level of incorporation, a company licensed in several states may face examiners working to different depths of AI review in the same year. No comment period and no compliance deadline attach to the release, and the material carries no independent legal force until a state agency adopts it into its examination programme.
Licentium advises financial services firms and technology vendors on the regulation of AI in licensed activity. Work we undertake includes building AI use case inventories, preparing risk tiering documentation, reviewing third-party AI supply contracts, and examination readiness across multi-state licensing.