GovernanceCore
Trustworthiness

Transparency

Working definition

Reviewed 30 July 2026

Transparency is the appropriate disclosure of information about an AI system’s existence, purpose, operation, data, capabilities, limitations, governance, and impacts. What must be disclosed—and to whom—depends on role, risk, law, and the decision the information must support.

Context

Why it matters

Developers, deployers, reviewers, affected people, and regulators need different information to evaluate, operate, challenge, and oversee AI responsibly.

Operating note

What this looks like in practice

  1. 01Map audiences to the decisions they need to make and the evidence required.
  2. 02Disclose AI interaction, material limitations, uncertainty, and recourse where relevant.
  3. 03Protect security, privacy, and intellectual property without using them as blanket reasons for opacity.

Sources & further research

Primary authority anchors the definition. Research links add conceptual or operational depth. External sources may update independently; always verify legal duties against the current official text.

  1. Official source

    The Language of Trustworthy AI: An In-Depth Glossary of Terms

    Provides a shared vocabulary for transparency, explainability, interpretability, and accountability.

    NIST
    2023    Open source ↗
  2. Official source

    Regulation (EU) 2024/1689, Articles 13, 50, and 53

    Creates different transparency duties for high-risk systems, certain AI interactions, and GPAI models.

    European Union
    2024    Open source ↗