AI Glossary

Responsible AI

Also called trustworthy AI

What does “Responsible AI” mean?

Responsible AI: Umbrella term for practices meant to make AI systems safe, fair, transparent, accountable and respectful of privacy throughout their life cycle. In organisations it usually means concrete processes: risk reviews, documentation, testing for bias, human oversight and ways to report problems. Critics note it can become a label without substance if not backed by real checks.

Responsible AI visual guide: Umbrella term for practices meant to make AI systems safe, fair, transparent, accountable and respectful of privacy throughout their life cycle. In organisations it usually means concrete processes: risk reviews, documentation, testing for bias, human oversight and ways to report problems. Critics note it can become a label without substance if not backed by real checks.
Visual guide Open full size (new tab)

What is an example of “Responsible AI”?

A bank might require a fairness audit, a model card and sign-off from a review board before any AI model affects lending decisions.

Keep learning

See also

AI safetyTermFairnessTermModel cardTermAI regulationTerm

Same topic

AI biasTermData privacyTermEU AI ActTermExplainabilityTermGuardrailsTerm