OCC's Gould says AI will not be a separate supervisory risk category
For the chief risk officer at an OCC-supervised bank
Comptroller of the Currency Jonathan Gould said in remarks to the Exchequer Club on Nov. 18 that the OCC is not treating banks' use of what his text calls "Super Intelligence" (SI) as a separate risk category, according to the OCC's published remarks; Banking Dive reported the same point as AI.21
Gould said the OCC issued a supervisory framework this month for banks' use of SI, and that it helps examiners evaluate that use within established risk management practices.2 He said SI can help banks detect fraud and serve customers faster, and can make sophisticated capabilities more accessible to smaller institutions.2
Gould said the OCC treats SI as "a component of preexisting business operations and practices," just as banks are not required to establish a separate SI risk appetite.2
He said: "Supervisory judgment should reflect what new technology does and the potential consequences of its use."2
Gould said the OCC and the FDIC finalized a uniform definition of unsafe or unsound practices and set clearer standards for Matters Requiring Attention, so banks understand the basis for a criticism and what resolves it.2
He said examiner judgment will be bolstered by eliminating checklists and refocusing on material financial risk; the remarks do not say which checklists are affected or when.2
If examiners judge AI use inside existing model, vendor and operational risk reviews, banks cannot count on an AI-specific rulebook. A documented inventory that maps each AI use to an existing risk framework is likely to be the main evidence examiners see.
Gould named capital requirements, Bank Secrecy Act and anti-money laundering reforms, and Community Reinvestment Act reforms as short- to medium-term priorities, without giving dates.2
Our take. Gould says examiners will judge AI as part of existing operations and are dropping checklists for judgment. A standalone AI risk program would answer a question the OCC says it won't ask; your model and vendor reviews get asked instead.
Action. Model risk: Tag each AI use case in the inventory to an owning model, third-party or operational risk review before the next exam.
What our records say
The OCC supervises AI at national banks through model risk management, safety-and-soundness examination and a dedicated examiner framework, not through AI-specific rules, and OCC Bulletin 2026-13, issued with SR 26-2, is the main AI-relevant requirement.12 OCC Bulletin 2026-13 requires risk-based, materiality-driven model risk management: sound development and testing, independent validation with outcomes analysis, effective challenge, and board and senior-management governance.3
From our records
- OCCregulator · binding Supervisory guidance · latest OCC Bulletin 2026-54 (Nov 2026), OCC Semiannual Risk Perspective, Spring 2026 (May 2026)
- FDICregulator · binding Supervisory guidance · latest Hill House oversight testimony (Jun 2026) (Jun 2026), FDIC FIL-15-2026 (Apr 2026)
Sources
- Report6 highlights of the OCC’s plan for AI, charters and community banks | Banking Dive Banking Dive
- Primary sourcehttps://occ.gov/news-issuances/news-releases/2026/nr-occ-2026-102a.pdf OCC
- Also reported byOCC Head Says Reforms Unlock $64 Billion in Capital for Community Banks PYMNTS
- Also reported byOCC’s Gould defends charter approvals American Banker