Local-first evidence workspace · no legal verdict
Make AI transparency obligations visible as operational work
Review six practical Article 50 transparency areas, assign owners, record evidence references and export a lightweight audit trail. Data stays in this browser session unless you export it.
Article 50 of the EU AI Act applies from 2 August 2026, but whether a specific obligation applies depends on the system, provider/deployer role and use case. This workspace organises evidence only. It does not determine legal scope, certify compliance or replace legal review.
A narrow transition applies only to Article 50(2) machine-readable marking and detection for AI systems placed on the market before 2 August 2026: providers of those systems must comply with that obligation from 2 December 2026. The other Article 50 duties are not postponed.
Applicable
6
Ready
0
Gaps
0
Unreviewed
6
Direct AI interaction disclosure
Systems intended to interact directly with people
Record how people are informed that they are interacting with an AI system, unless this is obvious from the circumstances.
Machine-readable synthetic-content marking
Providers of systems generating synthetic audio, image, video or text
Record the technical method used to mark generated or manipulated content in a machine-readable format where Article 50(2) applies. If the AI system was placed on the market before 2 August 2026, check the Commission's limited transition for this marking/detection obligation: providers of those systems must comply from 2 December 2026. This transition does not extend the other Article 50 obligations.
Limited Article 50(2) transition · 2 December 2026
This applies only to the marking and detection obligation for systems placed on the market before 2 August 2026. Content generated before 2 August 2026 does not have to be labelled retroactively.
Emotion / biometric categorisation notice
Deployers of emotion-recognition or biometric-categorisation systems
Record the notice shown to exposed people and the operational owner responsible for it.
Deepfake disclosure
Deployers generating or manipulating deepfake image, audio or video
Record the disclosure that content has been artificially generated or manipulated and where that disclosure appears.
AI-generated public-interest text disclosure
Deployers publishing AI-generated or manipulated text on matters of public interest
Record the disclosure workflow or the human-review/editorial-control basis relied on for the publication process.
Clear, accessible and timely disclosure
Applicable Article 50 disclosures
Record how disclosures are presented clearly, accessibly and at the latest at the first interaction or exposure, as applicable.