Login Get a Demo
01 Glossary · Governance and risk

What Is Model Risk Management?

How banks govern the models behind their decisions, what US regulators changed in April 2026, and where generative and agentic AI now sit.

The short answer

What is model risk management?

Last updated
Read time
4 min read
From
Cranium AI
Also called
MRM
Share on LinkedIn

Model risk management is the set of governance, development, validation and monitoring practices an organization uses to control the harm that wrong or misused models can cause.

It grew out of banking supervision and covers every model that informs a decision, built in-house or bought from a vendor, for as long as that model is in use.

Key takeaways
  • US regulators define model risk as the potential for adverse financial consequences from decisions based on model output.
  • The Fed, OCC and FDIC replaced SR 11‑7 with revised guidance on April 17, 2026 (SR 26‑2, OCC Bulletin 2026‑13).
  • The revised guidance states that generative and agentic AI models are not within its scope.
  • The UK’s PRA sets five principles in SS1/23, covering in-house and vendor models alike.
By the numbers Each figure links to its sourceAll sources 
02 The rulebook

What do regulators mean by model risk?

Two texts anchor most programs today, and their wording decides what lands in your inventory.

US revised guidance (SR 26‑2)UK PRA SS1/23
Issued byFederal Reserve, OCC and FDICPrudential Regulation Authority
In effectApril 17, 2026, replacing SR 11‑7May 17, 2024; current version April 23, 2026
A model isA complex quantitative method, system or approach that applies statistical, economic or financial theories to produce quantitative estimatesA quantitative method, system or approach that applies statistical, economic, financial or mathematical theories, techniques and assumptions
On AINon-generative, non-agentic AI models are covered. Generative and agentic AI models are notAddresses AI techniques such as machine learning, as far as they apply to models generally
Written forBanking organizations, most relevant above $30 billion in total assetsUK banks, building societies and designated investment firms with internal model approval
03 What it contains

What does a model risk program contain?

The US guidance covers development and use, validation and monitoring, governance and controls, and vendor products. In practice that means four working parts.

01

Model inventory

A record of every model in use, with enough information to understand its risks. Materiality, set by a model’s purpose and exposure, decides how much scrutiny it gets.

AI inventory
02

Development and use

Documented design, data and assumptions, plus users who understand a model’s limitations and monitor its performance.

03

Validation and monitoring

An evaluation of whether each model performs as expected, generally before first use and again as the model changes.

Model validation
04

Effective challenge

Critical analysis by objective experts with the standing to bring about change. SR 11‑7 called it a guiding principle of model risk management.

04 GenAI gap

Where do generative and agentic AI fit?

This is the question model risk teams are working through in 2026.

01

Out of scope, for now

A footnote in the revised guidance calls generative and agentic AI models novel and rapidly evolving, and not within its scope. A bank’s own governance should set controls for what it does not cover.

Read SR 26‑2
02

More to come

The agencies plan a request for information that will consider banks’ use of AI, including generative and agentic AI.

OCC Bulletin 2026‑13
03

Vendor models stay in

Proprietary code is no exemption. The guidance expects vendor products to be validated, by internal or outside parties, and monitored for fitness for purpose.

04

Where Cranium AI fits

Cranium AI is not a validation workbench and ships no banking guidance as a built-in framework. Its AI‑BOMs, AI Cards, test results and runtime records support a bank’s own model risk process.

Banking model risk management
05 Keep going

Related reading.

06 Questions

Frequently asked questions.

Straight answers, kept current. Last updated .

Ask us anything →
Did SR 26‑2 replace SR 11‑7?

Yes. On April 17, 2026 the Federal Reserve, OCC and FDIC issued revised guidance on model risk management. The Federal Reserve’s letter, SR 26‑2, supersedes SR 11‑7 from April 2011, and OCC Bulletin 2026‑13 rescinds OCC Bulletin 2011‑12. The OCC says the new guidance does not set enforceable standards or prescriptive requirements.

Does model risk management apply to generative AI?

Under the US guidance, not directly. It states that generative and agentic AI models are not within its scope, while non-generative, non-agentic AI models are. The risk remains: a bank’s own risk management and governance practices should set controls for tools the guidance does not cover.

Is model risk management only for banks?

The formal guidance is written for banks. The US text is most relevant to banking organizations with more than $30 billion in total assets. The practices travel well: any organization that relies on models can keep an inventory, validate independently and monitor. For AI more broadly, see AI risk management.

07 See it for yourself

From the definition
to the evidence.

See how Cranium AI discovers, observes, governs, secures and proves every AI model, agent and vendor in your enterprise.