What is explainable AI (XAI)?
Explainability helps us understand why a system suggested a particular prediction or decision.
What is it?
Explainable AI (XAI) is the set of methods that help us understand why a system suggested a decision: which data influenced it most and what would change it. An explanation has to be understandable and also faithful to what really happens inside the model.
An everyday example
When a bank turns down a loan, it should be able to tell you which factors decided it, such as income or existing debt, not just “the system calculated it”.
Why does it matter?
Without an explanation it is hard to find errors, bias or misuse. For decisions about people, explainability is often a legal requirement as well.
A common misconception
That a model’s own explanation is enough. A chatbot can write a convincing explanation that does not match the real reason for its answer.
In more detail
What counts as an explanation?
An account of the influential data, rules or reasons that can be checked.
Is the model’s own explanation enough?
No. A well-written explanation does not prove that it is the real reason for the output.
When does it matter most?
When the system affects health, employment, finances or rights.
What is the difference between transparency and explainability?
Transparency tells you how a system is designed and used; explainability tries to account for a specific behaviour.
Remember: A good explanation has to be checkable, not just easy to understand.
Reliable sources
Last reviewed 1 November 2026. An explainer, not professional, legal or medical advice. For rules and figures that change, check the official version.