Model risk management training: SR 26-2, validation and AI models
A practical course for teams that build, validate, use or oversee models. It covers the SR 26-2 risk-based approach, building a model inventory, materiality tiering, effective challenge, validation of conceptual soundness and outcomes, ongoing monitoring, vendor models, and governance for generative and agentic AI that sits outside SR 26-2.
Model risk management changed in April 2026. This course brings teams up to date and gives them working templates.
Who is this course for?
Model risk managers, validators, quantitative analysts, data scientists, second-line risk staff and internal auditors.
What does the course cover?
- SR 26-2 compared with SR 11-7
- Model definition, inventory and materiality tiering
- Validation: conceptual soundness, outcomes analysis, monitoring
- Vendor and third-party models
- Governance for generative and agentic AI
- Hands-on exercise: tier and plan validation for a sample inventory
How is it delivered?
In-house, at a venue or online, as a half-day board session or a one- to two-day programme for specialists. Content is adapted to your institution's jurisdiction and AI use, using the Doorstep risk register for practical exercises. Led by Ahmer Nehal, former Head Risk and Fraud and SVP Operational Risk, who has trained senior management in Pakistan and the UAE.
Primary sources (reviewed 17 Oct 2026)
- Federal Reserve, OCC and FDIC: SR 26-2, Revised Guidance on Model Risk Management (17 April 2026)
- Financial Stability Board: Sound Practices for Responsible Adoption of Artificial Intelligence, final report (15 October 2026)
This page explains supervisory guidance in plain language. It is not legal advice; confirm how each text applies to your institution and jurisdiction.
Frequently asked questions
Is the course only for US banks?
No. SR 26-2 is used as an international benchmark and the course covers FSB and MAS expectations as well.
Related pages
Put these controls to work this month
Start on the cloud risk management software within 24 hours, or ask for training or advice tailored to your institution.