AI Transparency: What You Have to Disclose

AI transparency used to be a nice-to-have. Under the EU AI Act it is increasingly a duty — and one that is straightforward to meet if you plan for it.

Two kinds of transparency

• User-facing — telling people when they are interacting with an AI system, or when content is AI-generated.
• System-facing — documentation, model cards and records that let auditors and deployers understand how a system works.

Make it routine

Bake disclosure into your product and documentation process so transparency is produced as a by-product of building, not a scramble at review time.

Where this fits

This post connects the EU AI Act and AI governance pillars. Transparency is a core pillar of responsible AI.