Which of the four transparency duties apply to each AI system you provide or deploy — with the legal basis, a reperformable assessment trail, the applicable deadline and a residual-risk view. The Digital Omnibus moved the high-risk regime; it did not move Article 50.
The Digital Omnibus amendments were endorsed by the European Parliament on 16 June 2026 and approved by the Council on 29 June 2026; publication in the Official Journal was still pending, so reconcile the 2 December 2026 transitional for legacy Art. 50(2) marking against the final OJ text. The Commission's Article 50 guidelines (8 May 2026) remain in draft. This tool covers Article 50 only — Article 5 prohibitions, the high-risk regime (Chapter III) and GPAI obligations are out of scope. It is an assessment aid, not legal advice.
Add each AI system and set your role — provider, deployer or both. Role determines which duties are yours and which you verify.
Work through the five duty blocks. Every question carries its article reference; "unsure" is treated conservatively and flagged.
Read the conclusion, the full answer trail behind it, the deadline and the residual risk. Export to Excel or print the memo.
Browser-only · nothing you enter leaves this page · no storage — export JSON to keep your work
Paraphrased for assessment; the article reference on every question and conclusion is the reconciliation hook. Reperform against the official text on EUR-Lex.
Art. 50(1) Provider
Systems intended to interact directly with natural persons must be designed and developed so those persons are informed they are interacting with AI — unless that is obvious to a reasonably well-informed, observant and circumspect person in the context of use. Exception for legally authorised law-enforcement use (unless the system is open to the public to report offences).
Art. 50(2) Provider
Providers of systems (including GPAI) generating synthetic audio, image, video or text must mark outputs in a machine-readable format, detectable as artificially generated or manipulated — effective, interoperable, robust and reliable so far as technically feasible. Exceptions: assistive/standard-editing functions that don't substantially alter the deployer's input or its meaning; authorised law-enforcement use. Transitional: systems on the market before 2 Aug 2026 have until 2 Dec 2026 (Digital Omnibus).
Art. 50(3) Deployer
Deployers must inform exposed persons of the system's operation and process personal data in line with EU data-protection law. Exception for legally permitted law-enforcement use with safeguards. Check Article 5 first — several uses (e.g. emotion recognition in workplaces and education) are prohibited outright.
Art. 50(4) §1 Deployer
Deployers generating or manipulating image, audio or video that constitutes a deep fake (Art. 3(60)) must disclose the artificial generation or manipulation. Where the content is part of an evidently artistic, creative, satirical, fictional or analogous work, disclosure is limited to an appropriate form that does not hamper display or enjoyment.
Art. 50(4) §2 Deployer
Deployers publishing AI-generated or manipulated text to inform the public on matters of public interest must disclose it — unless the content has undergone human review or editorial control and a natural or legal person holds editorial responsibility for its publication (or a law-enforcement authorisation applies).
Art. 50(5)–(6)
Information under 50(1)–(4) must be clear and distinguishable, given at the latest at the first interaction or exposure, and conform to applicable accessibility requirements. These duties sit alongside — not instead of — the high-risk regime (Chapter III) and other EU or national transparency law (e.g. GDPR, DSA).
| Likelihood \ Impact | Low | Medium | High |
|---|---|---|---|
| Low | Low | Low | Medium |
| Medium | Low | Medium | High |
| High | Medium | High | High |
This is an informational self-assessment aid built by a risk practitioner, not legal advice, and it does not create a lawyer–client relationship. Classifications depend on the accuracy of your answers and on interpretations that the Commission's final guidelines may refine. Validate conclusions against the official texts and your own counsel before relying on them.