For Heads-of · Practitioner
Counterfactual Explanation
An explanation of an automated decision that states the smallest change to the inputs that would have changed the outcome, such as an application that would have been approved with a higher declared income. It tells the affected person what would have made the difference without exposing the model.
- explainability
- counterfactuals
- right to explanation
- automated decisions
Why it is useful
Most people affected by a decision want to know why, and what would change it, not how the model works. A counterfactual answers both in their terms and supports the right to contest. It needs no access to the model's internals, which makes it practical for third-party and opaque systems. Wachter, Mittelstadt and Russell proposed it in 2017 as a way to meet GDPR transparency without disclosing the algorithm.
Its limits
A counterfactual can point to something the person cannot change, such as age, or to a proxy for a protected characteristic, which reveals discrimination rather than explaining a decision. Several different counterfactuals may be valid, and the one chosen shapes the message. It explains one decision, not the system's overall behaviour.
Where it sits
It is one technique under explainability, alongside feature attribution, and it is what makes contestability practical. The register entry it addresses is inadequate explainability for affected individuals; automated decision-making safeguards is the control.
Sources
Related Terms
Automated Decision-Making
A decision about a person taken by technological means without meaningful human involvement. In UK data protection law, Articles 22A to 22D of the UK GDPR (inserted by the Data (Use and Access) Act 2025 and in force since 5 February 2026) allow significant automated decisions, but only with safeguards: information, the right to make representations, human intervention and the right to contest.
Contestability
The ability of a person affected by an AI-assisted decision to challenge it and have it reconsidered by someone with the authority to change it. It is one of the safeguards UK data protection law now attaches to significant automated decisions.
Explainability
The ability to give a person an account of why a system produced a particular output, in terms they can act on. Explainability is a property of the account and its audience, not of the model: an opaque model can be explainable, and an interpretable one can fail to be explained.
Feature Attribution
Methods that estimate how much each input contributed to a model's prediction, such as SHAP and LIME. They are standard in model validation and explanation, but they approximate the model rather than describe it, and they can be manipulated.