AI Governance Engineer
GitHub

Explainability

By Updated

In NIST's framing, a representation of the mechanisms behind a system's operation: how a decision was made 1. In practice a per-decision explanation such as feature attributions, reason codes or a counterfactual.

Developed in
ch. 16, Transparency, interpretability and explainability
ch. 11, Contrast pairs
Chapters
ch. 11, AI Defined · ch. 16, Fairness & XAI
Contrast with
Transparency · Interpretability
Source
1 numbered reference, listed below

Commonly confused

  • Transparency versus Explainability

    Transparency
    In NIST's framing, how far information about an AI system and its outputs reaches the people who interact with it: what happened.
    Explainability
    In NIST's framing, a representation of the mechanisms behind a system's operation: how a decision was made.

    The differenceWhat happened, from records of what ran, against how one decision was made.

    Why it mattersTwo artefacts: registry, cards and logs for the first; an explanation method with a fidelity test for the second.

Where it is used

8 chapters of the Body of Knowledge use the term. Each link opens the first section that does.

  • Concept drift

    A change in the relationship between a system's inputs and the correct output, so the same input should now get a different answer.

  • Data drift

    A change in the distribution of the inputs a system sees in production relative to the data it was validated on, such as a new customer segment or a changed…

  • Opacity

    The inability of a person to follow how a system reached an output.

  • OECD AI Principles

    The five values-based principles (inclusive growth and well-being; human rights, fairness and privacy; transparency and explainability; robustness, security…

  • Reason code

    A stable, human-readable statement of a principal factor behind an adverse decision, mapped from the factors the model actually scored and versioned with the…

Sources

  1. [1] Artificial Intelligence Risk Management Framework (AI RMF 1.0), NIST AI 100-1 (risk tolerance and residual risk; seven trustworthy characteristics; transparency answers "what happened", explainability "how", interpretability "why"; MAP 1.1 intended purposes; MANAGE 1.1 go/no-go determination; profiles). NIST. 2023-01-26. https://doi.org/10.6028/NIST.AI.100-1 (verified: primary)

Definitions of legal terms paraphrase the cited text, which governs. Dated statements are as of .

Cite this term

García Aibar, J. (2026). Explainability. In AI Governance Engineering: The Thesis & Body of Knowledge (v0.5.0), Glossary. https://doi.org/10.5281/zenodo.22956197. https://aigovernanceengineer.com/glossary/explainability. CC BY 4.0

BibTeX

@misc{aige2026explainability,
  author  = {Jorge García Aibar},
  title   = {{Explainability}},
  note    = {Glossary, AI Governance Engineering: The Thesis \& Body of Knowledge, version 0.5.0},
  year    = {2026},
  doi     = {10.5281/zenodo.22956197},
  url     = {https://aigovernanceengineer.com/glossary/explainability}
}