Lab 002 · LIVE · Last verified 19 August 2026

AI Transparency Checker

A structured starting point for product teams—not a legal verdict.

Product review questionsDoes a person interact directly with AI?Does the system generate or modify text, images, audio or video?Could the output be mistaken for human-created content?Can the system take consequential actions?Is provenance or an action history currently visible?

Questions to review

Select the relevant product behaviors

Make AI involvement legible when it materially changes the interaction.
Review provenance, user control, limitations and reversibility.
Map provider/deployer roles and relevant dates with qualified counsel.
Keep the decision logic and product action history auditable.

Educational information only — not legal advice and not a compliance determination.

How this works

The checker surfaces product questions around identity, provenance, control, limits and consequence. It never outputs compliant or non-compliant.

FAQ

Is this legal advice?

No. Confirm roles, scope and deadlines with qualified counsel.

Primary Sources

EU AI Act Service Desk — enforcement timeline ↗European Commission — GPAI provider guidelines ↗Related Journal: AI Act Enforcement Has Started