Glossary
AI Transparency
The availability of meaningful information about an AI system’s existence, purpose, operation, limitations and governance.
Last reviewed 2026-08-22
In plain language
Transparency ranges from user-facing notices (“you are interacting with an AI system”) to technical documentation, training-data summaries, public registers and explainability of individual decisions. More disclosure is not automatically more useful; poorly designed transparency can overwhelm people or reveal attack surfaces. Legal transparency duties in data-protection and AI-specific laws are narrower than the rhetorical use of the word.
Why it matters for AI governance
Transparency is a prerequisite for contestability, journalism and democratic oversight, but it is not a substitute for limits on harmful uses.