VITNAcontainment for rogue AI agents
Try the demoSign in
[ eu ai act · article 50 · applies 2 august 2026 ]

Which Article 50 obligations
apply to you?

The EU AI Act's transparency obligations apply from 2 August 2026 (machine-readable marking for systems placed on the market before that date: by 2 December 2026), and they reach any company whose output is used in the EU — established there or not. A few questions tell you which of the five obligations appear to apply to you. No signup. When you finish, we count which obligations your answers triggered, with no email address, IP address or account attached. Your email address is stored only if you ask us to email you the result. The result is a shareable link either way.

This tool determines what applies. It never claims to determine whether you have met it, and it does not score you.

[ scope ]

Do people in the EU use your product, does its output reach people in the EU, or are you established or located in the EU and using AI systems in your own operations?

[ the five obligations, in brief ]
[ Article 50(1) · provider ]

Tell people they are talking to an AI

Systems that interact directly with people must be designed so a person knows they are interacting with an AI.

[ Article 50(2) · provider ]

Mark synthetic output so machines can detect it

Generated audio, images, video and text must be marked in a machine-readable format, for example by watermarking, metadata or content credentials. Applies from 2 August 2026 to systems placed on the EU market from that date (no grace period); systems placed on the EU market before then have until 2 December 2026 under the Digital Omnibus on AI. Placed on the EU market before 2 August 2026? Your marking deadline is 2 December 2026 — check it at /ai-act-december-2026.

[ Article 50(3) · deployer ]

Tell people when emotion or biometric categorisation is used

People exposed to emotion recognition or biometric categorisation must be informed of its operation.

[ Article 50(4), first subparagraph · deployer ]

Disclose deepfakes

AI-generated or manipulated image, audio or video content that resembles existing people, objects, places, entities or events and would falsely appear authentic or truthful must be disclosed, at the latest when a person is first exposed to it.

[ Article 50(4), second subparagraph · deployer ]

Disclose AI-generated public-interest text

AI-generated or manipulated text published to inform the public on matters of public interest must be disclosed, unless the text has undergone human review or editorial control and a natural or legal person holds editorial responsibility for its publication.

Penalties for Article 50 violations reach EUR 15,000,000 or 3% of total worldwide annual turnover, whichever is higher (Article 99(4)); for SMEs, including start-ups, and small mid-cap enterprises (SMCs), the cap is whichever of the two is lower (Article 99(6) and (6a)). This page covers Article 50 only — the prohibited practices in Article 5, high-risk obligations under Annex I and Annex III, and general-purpose AI model obligations are separate assessments with separate dates.