The Federal Reserve supervises bank AI through its model risk and safety-and-soundness framework rather than AI-specific rules. Its landmark document, SR 11-7 (2011), was superseded in April 2026 by revised interagency model risk management guidance issued jointly with the OCC and FDIC — which notably excludes generative and agentic AI from its scope. Fed leadership in 2026 has publicly favored an innovation-permissive stance, emphasizing that supervisors should not impede responsible AI adoption.
| Full name | Board of Governors of the Federal Reserve System |
| Role | Central bank and prudential supervisor |
| Force on banks | Supervisory guidance |
| Applies to | State member banks, bank and savings-and-loan holding companies, and US operations of foreign banks |
| Key document | SR 26-2 — Revised interagency Model Risk Management guidance (Apr 2026, supersedes SR 11-7) |
| Latest move | Apr 2026 revised model risk guidance; May 2026 Vice Chair speech on AI in the financial system |
SR 11-7 was arguably the most influential model-governance document in world banking — the template for how banks validated any quantitative model, including ML. Its April 2026 replacement keeps the core disciplines (validation, effective challenge, governance) while modernizing for machine learning and deliberately leaving generative and agentic AI to enterprise risk management pending further guidance.
Vice Chair for Supervision remarks in May 2026 framed AI as a technology the financial system should adopt with appropriate controls, positioning the Fed against pre-emptive AI-specific rulemaking. Banks should expect examination focus on governance, third-party (foundation-model vendor) risk, and data controls rather than new AI rules.
| Date | Type | Document / event |
|---|---|---|
| Apr 4, 2011 | Guidance | SR 11-7: Supervisory Guidance on Model Risk Management. The foundational model-governance framework used globally for bank models, including early ML. Superseded April 2026. |
| Mar 31, 2021 | Consultation | Interagency RFI on AI in financial services. Joint request for information with OCC, FDIC, CFPB, and NCUA on bank use of AI and machine learning. |
| Apr 17, 2026 | Guidance | SR 26-2: Revised interagency Model Risk Management guidance. Fed, OCC, and FDIC jointly supersede SR 11-7 with a principles-based framework, most relevant to banks over $30B in assets; generative and agentic AI are excluded from scope as novel and rapidly evolving. |
| May 1, 2026 | Speech | Vice Chair for Supervision speech on AI in the financial system. Sets out the Fed's supervisory philosophy on AI adoption in banking. |
- Whether the Fed issues follow-on guidance covering generative and agentic AI, which the 2026 model risk revision deliberately left out
- Examination practice at large banks: how third-party foundation-model dependence is treated under existing vendor-risk guidance
- FSOC's annual report treatment of AI as a systemic vulnerability
Is SR 11-7 still in effect?
No. On April 17, 2026 the Federal Reserve, OCC, and FDIC issued revised interagency model risk management guidance that supersedes the 2011 SR 11-7 framework. The revision covers traditional and ML models but explicitly excludes generative and agentic AI.
Does the Federal Reserve have AI-specific rules for banks?
No. The Fed regulates AI through existing frameworks — model risk management, safety and soundness, and third-party risk guidance. Its 2026 public posture favors allowing responsible AI adoption rather than imposing AI-specific rulemaking.
How should banks govern generative AI if it's excluded from model risk guidance?
The 2026 interagency guidance directs banks to apply their broader risk-management and governance practices to generative and agentic AI — meaning enterprise risk frameworks, vendor risk management, data governance, and board oversight, rather than formal model-validation requirements.
Follow every move these regulators make
6 curated AI stories for banking executives · Every morning · Free
Subscribe to BankingNewsAI →