Transparency
AI Interaction and Biometric Transparency Notices
What this control does
Provide clear, timely and accessible notices when people interact with AI or are exposed to qualifying emotion-recognition or biometric-categorisation systems.
How to implement
Use this control for the notices required by the relevant Article 50 route; first confirm the underlying use is permitted.
- Identify direct AI interactions and qualifying emotion-recognition or biometric-categorization uses. Separate the provider's design duty from the deployer's notice duty.
- Draft a clear notice and place it where people encounter the system, no later than the required first interaction or exposure.
- Check accessibility and visibility across the actual channels and devices. Record the reasoning for any relied-on exception.
- Review the notice after material interface or use changes, and correct missing or confusing disclosures. A notice does not make an otherwise prohibited use lawful or replace separate personal-data requirements.
Suggested timing and triggers
At the first applicable interaction or exposure; before release; after material interface, channel or use changes.
Evidence examples
Notice applicability and exception assessment Approved notice text and interface placement Screenshots or recordings from representative user journeys Accessibility and timing checks Deployment approval and revision history
How to check this control
Walk through the experience as a first-time user on relevant channels. Confirm that the notice appears at the required point and is distinguishable and accessible. Compare the live experience with the approved wording and any exception reasoning.