# Building an AI Governance Framework for a Bank

> Financial Services · ibl.ai
> Source: https://ibl.ai/courses/financial-services/ai-governance-framework-for-banks
> Last updated: 2026-11-18

*Write the AI governance framework a bank needs: board oversight, an inventory of every AI use, risk tiering, model risk and third-party oversight, fair lending review and evidence an examiner can follow*

**A financial services course from ibl.ai (self-paced, 7 to 9 hours, free).** A hands-on, self-paced course for bank and credit union risk, compliance, model risk and technology leaders. Build an AI governance framework from current guidance: SR 26-2 (April 2026) and its generative and agentic AI carve-out, the interagency third-party guidance and its 2026 proposed replacement, Regulation B adverse action reasons, Treasury's AI cybersecurity report and the NIST AI RMF, ending in a framework and a twelve-month roadmap.

- **Price:** Free
- **Format:** Self-paced, online
- **Time:** 7 to 9 hours
- **Pace:** 3 hours per week
- **Level:** Advanced
- **Language:** English
- **Subject:** Risk Management

[Enroll](https://lms.ibl.ai/platform/financialservices/courses/course-v1:financialservices+ai-governance-framework-for-banks+self-paced)

[Course outline and sources](https://ibl.ai/solutions/financial-services/course/ai-governance-framework-for-banks)

## Course description

Building an AI Governance Framework for a Bank teaches risk, compliance, model risk, technology and internal audit leaders to write the framework that governs every AI use at a bank or credit union. It starts from the guidance as it stands on 18 November 2026: SR 26-2, issued by the Federal Reserve, OCC and FDIC on 17 April 2026, which superseded SR 11-7 and SR 21-8, is expected to be most relevant to banking organizations with over $30 billion in total assets, and places generative and agentic AI outside its scope; the 2023 interagency third-party guidance, which remains in force while a September 2026 proposal to replace it is considered; Regulation B's requirement of specific principal reasons for adverse action, after the CFPB withdrew its adverse action circulars in May 2025; Treasury's March 2024 report on AI-specific cybersecurity risks; and the NIST AI RMF and its Generative AI Profile. Through five labs and five templates you build a policy gap map, a committee charter, an inventory that finds AI inside vendor products, a validation and monitoring plan, AI vendor due diligence questions, an adverse action review, a risk-to-control map, and finally a framework document and twelve-month roadmap. The course teaches how to build and evaluate a framework; your compliance, legal and model risk functions and your examiners decide whether it meets any requirement.

### About this course

Banks are adopting AI faster than their policies describe it, and much of it arrives inside vendor products. This course builds an AI governance framework from the supervisory guidance and regulation that exist today, read from the agencies' own pages: [SR 26-2, Revised Guidance on Model Risk Management](https://www.federalreserve.gov/supervisionreg/srletters/SR2602.htm) (Federal Reserve, OCC and FDIC, 17 April 2026), which superseded SR 11-7 and SR 21-8 and places generative and agentic AI outside its scope; the 2023 Interagency Guidance on Third-Party Relationships and the September 2026 proposal to replace it; Regulation B on adverse action notices, and the CFPB's 2025 withdrawal of its adverse action circulars; Treasury's report on AI-specific cybersecurity risks in financial services; and the NIST AI Risk Management Framework. You leave with a framework document and a twelve-month roadmap. The course teaches how to build and evaluate a framework; whether it meets any legal or supervisory requirement is for your compliance, legal and model risk functions and your examiners to decide.

### What you will be able to do

- Define the scope of an AI governance framework for a bank, including a policy definition of AI that handles the SR 26-2 model definition and its exclusion of generative and agentic AI, and connect it to model risk, third-party risk and enterprise risk programs.
- Assign AI risk ownership from the board to the first line: board reporting, an AI risk committee charter and a responsibility matrix with named owners and effective challenge.
- Inventory and tier every AI use, including AI embedded in vendor products and tools staff adopt on their own, with fields that support risk management at the individual and aggregate levels.
- Apply SR 26-2 to AI models in its scope, including vendor models, and specify equivalent testing, monitoring and switch-off controls for generative and agentic AI using the NIST AI RMF and its Generative AI Profile.
- Oversee AI vendors across the third-party life cycle in force, and specify how AI used in credit decisions produces the specific principal reasons Regulation B requires, routing every conclusion to compliance.
- Map AI-specific security risks to existing controls, specify data location, access and logging an examiner can follow, and write the framework and a twelve-month roadmap a risk committee can adopt.

### Who this course is for

**Roles:** chief risk officers and heads of model risk management, compliance and legal leaders, heads of AI, data and technology, and internal audit leaders.

**Institutions:** banks and credit unions of any size; SR 26-2 is expected to be most relevant to banking organizations with over $30 billion in total assets, and the course shows how to tailor for smaller ones. The framework method also suits smaller financial firms such as asset managers and hedge funds, whose compliance teams map it to the rules that apply to them. The examples follow a fictional regional bank.

**Before you start:** familiarity with your institution's model risk and third-party risk programs. No coding is needed.

### Format

**Time:** about 7 to 9 hours, self-paced. **Level:** advanced. **Hands-on:** five labs and five templates that build your framework piece by piece (a policy gap map, an AI inventory with tiering, a validation and monitoring plan, an adverse action review, a risk-to-control map), with an AI tutor that reviews each draft with questions.

**Structure:** a Start here section, four content sections of readings, worked examples, templates and labs, each ending in an AI assessment, and a capstone in which you write an AI governance framework and a twelve-month roadmap, followed by a reading list.

**Grading:** the four section AI assessments count for 60 percent of the grade and the capstone assessment for 40 percent. The pass mark is 70 percent.

### Currency

**Last verified 18 November 2026** against SR 26-2 and the Federal Reserve's 2026 SR letters list, OCC Bulletins 2026-13 and 2026-46, the 2023 interagency third-party guidance and the September 2026 proposal, Regulation B on eCFR, the CFPB withdrawal notice of 12 May 2025, the Treasury report of March 2024 and the NIST AI RMF pages.

The rules are still changing. The banking agencies have said they plan a request for information on model risk management and AI, including generative and agentic AI; the third-party proposal may be finalized in a different form; the CFPB says its withdrawals are not necessarily final; and NIST is revising the AI RMF. Check the current text before relying on any page or template.

## Outline

### Section 0 · Start here

- 0.1 Welcome and orientation

### Section 1 · Scope the framework and assign ownership

- 1.1 What the framework has to cover
- 1.2 Who owns AI risk
- 1.3 Section 1 assessment

### Section 2 · Inventory, tier and validate AI uses

- 2.1 Inventory and tiering
- 2.2 Model risk guidance after SR 26-2
- 2.3 Section 2 assessment

### Section 3 · Oversee third-party AI and explainable credit decisions

- 3.1 Third-party AI
- 3.2 Explainable credit decisions
- 3.3 Section 3 assessment

### Section 4 · Secure AI and write the framework

- 4.1 Securing AI systems and their data
- 4.2 Writing the framework and its first-year roadmap
- 4.3 Section 4 assessment

### Section 5 · Capstone and wrap-up

- 5.1 Capstone: An AI governance framework and twelve-month roadmap
- 5.2 Capstone assessment
- 5.3 Reading list and next steps

## Instructors

- ibl.ai Financial Services, Course team (ibl.ai)
