Being open about how an AI system works, what it can and cannot do, where its data comes from, and who is responsible for it.
Being open about how an AI system works, what it can and cannot do, where its data comes from, and who is responsible for it.
Transparency in AI operates at multiple levels: transparency about the AI system itself (capabilities, limitations, training data), transparency of the AI process (how decisions are made), and transparency of AI governance (who is accountable and how oversight is exercised). The EU AI Act mandates transparency through transparency obligations (Article 50) and technical documentation requirements (Articles 11–13). The OECD AI Principles include transparency and explainability as a foundational principle. Transparency is distinct from explainability: transparency concerns disclosure of information; explainability concerns interpretation of that information.
Organisations that proactively communicate AI transparency information — through model cards, system cards, and public documentation — build stakeholder trust and reduce the risk of regulatory challenges.
Like the nutrition labelling on food products — not every consumer will read or understand every detail, but the information must be present and accessible so those who want it can find it.
Being open about how an AI system works, what it can and cannot do, where its data comes from, and who is responsible for it.
Transparency in AI operates at multiple levels: transparency about the AI system itself (capabilities, limitations, training data), transparency of the AI process (how decisions are made), and transparency of AI governance (who is accountable and how oversight is exercised). The EU AI Act mandates transparency through transparency obligations (Article 50) and technical documentation requirements (Articles 11–13). The OECD AI Principles include transparency and explainability as a foundational principle. Transparency is distinct from explainability: transparency concerns disclosure of information; explainability concerns interpretation of that information.
Organisations that proactively communicate AI transparency information — through model cards, system cards, and public documentation — build stakeholder trust and reduce the risk of regulatory challenges.