Model Risk Management
Model risk management is the discipline of identifying, measuring, and controlling the risks a model poses to a business — that it's wrong, biased, misused, or drifts over time. It comes from regulated finance and now applies to AI: treat each model as a risk to be governed, not just a tool to be shipped.
Also known as: MRM
Model risk management (MRM) predates the AI boom — banks have governed statistical models this way for years under regulatory frameworks — and it transfers directly to AI. The premise is that any model in a consequential decision is a source of risk: it can be wrong, encode bias, be used outside its valid range, or degrade as the world changes. MRM is the structured practice of finding those risks, quantifying them, and putting controls around them.
In practice it pulls together much of this glossary under one governance umbrella: validation and evaluation before deployment, monitoring and drift detection in production, documented limitations and intended use, and clear ownership and sign-off. For regulated industries it’s often mandatory; for everyone else it’s the difference between deploying a model and being able to defend that decision when it matters.