Free tool · No signup · Runs in your browser

Free EU AI Act check: find your risk tier in about 3 minutes

Answer yes/no questions about one AI system. Get its risk tier and the obligations for your role, each with its article citation and date, under the AI Act as amended by Regulation (EU) 2026/1744.

By the Shieldra Compliance Team · Last updated July 2026

Key takeaways

  • The check covers every tier: prohibited, high-risk (Annex I and Annex III, including the Art. 6(3) derogation), transparency and minimal. Obligations are scoped to your role, because provider duties are not deployer duties.
  • Article 50 applies to AI systems that interact with people or generate content, whatever their risk tier. It is not limited to high-risk systems.
  • Article 50 sets five duties: tell people they are interacting with AI (50(1)), mark AI-generated content machine-readably (50(2)), inform people exposed to emotion recognition or biometric categorisation (50(3)), disclose deep fakes (50(4), first subparagraph), and disclose AI-generated text published to inform the public on matters of public interest (50(4), second subparagraph).
  • Breaching Article 50 can cost up to €15 million or 3% of worldwide annual turnover, whichever is higher (Art. 99(4)). For SMEs and start-ups (Art. 99(6)) and small mid-caps (Art. 99(6a)) the lower of the two applies.
  • Machine-readable marking has applied since 2 August 2026 for new systems. Providers of systems already on the market before that date have until 2 December 2026 (Art. 111(4)).
  • The high-risk regime was deferred by Regulation (EU) 2026/1744 to 2 December 2027 for Annex III systems and 2 August 2028 for Annex I products.
  • The same regulation added two prohibitions that apply from 2 December 2026 (non-consensual intimate imagery and child sexual abuse material) and rewrote the Article 4 AI-literacy duty as an obligation to take supporting measures.

What changed in 2026 (Digital Omnibus)

The Digital Omnibus on AI was adopted by Parliament on 16 June 2026 and by the Council on 29 June 2026, published as Regulation (EU) 2026/1744 on 24 July 2026, and has been in force since 27 July 2026. It deferred the high-risk regime: Annex III systems to 2 December 2027 and Annex I products to 2 August 2028. The dates are fixed; the standards-readiness trigger in the Commission proposal was dropped.

It added two prohibited practices to Article 5, systems that generate non-consensual intimate imagery and systems that generate child sexual abuse material, applying from 2 December 2026 at the €35 million / 7% penalty tier. It rewrote the Article 4 AI-literacy duty from ensuring a sufficient level of literacy to taking measures that support it, and extended the lower-fine rule to small mid-cap companies (Art. 99(6a)).

It kept the EU-database registration for providers that conclude an Annex III system is not high-risk under Article 6(3), with less information to file, and moved machinery into Section B of Annex I, where the AI requirements apply through the sector legislation (Art. 2(2)).

It did not defer Article 50. The transparency duties and the fines for breaching them have applied since 2 August 2026, as has the Commission's power to fine providers of general-purpose AI models (Art. 101). The only Article 50 relief: providers of generative systems already on the market before 2 August 2026 have until 2 December 2026 for machine-readable marking (Art. 111(4)).

The five Article 50 transparency duties

Tell people they are interacting with AI (Art. 50(1))

Providers must design the system so people are told they are interacting with AI, at the latest at the first interaction, unless that is obvious to a reasonably well-informed person. This covers chatbots, assistants, copilots and voice agents.

Mark AI-generated content machine-readably (Art. 50(2))

Providers of systems that generate audio, image, video or text must embed machine-readable markers so the output is detectable as AI-generated. A visible label alone does not meet this duty. The Commission's Code of Practice on marking and labelling AI-generated content (final version published 10 June 2026) is a voluntary way to show compliance.

Inform people exposed to emotion recognition or biometric categorisation (Art. 50(3))

Deployers must tell people that such a system is operating and process the personal data in line with the GDPR.

Disclose deep fakes (Art. 50(4), first subparagraph)

Deployers must disclose that image, audio or video content resembling real people, objects, places or events, which could falsely appear authentic, was generated or manipulated by AI. Evidently artistic or satirical works need only a disclosure that does not hamper the work.

Disclose AI-generated public-interest text (Art. 50(4), second subparagraph)

Deployers publishing AI-generated text to inform the public on matters of public interest must disclose it, unless the text went through human review or editorial control and someone holds editorial responsibility.

Who this applies to

The Act reaches AI systems placed on the EU market or whose output is used in the EU, so a company with no EU entity can still be in scope through its EU customers. Deployers established in the EU are covered wherever the output is used. Providers outside the EU of high-risk systems or general-purpose AI models must appoint an authorised representative in the EU (Arts. 22, 54).

Article 50 is broader than the high-risk regime. It attaches to ordinary products such as a support chatbot, an AI writing feature or a voice agent, not only to systems on the Annex III list.

Frequently asked questions

Does Article 50 only apply to high-risk AI systems?

No. Article 50 transparency duties attach to AI systems that interact with people or generate content, whatever their risk tier. The high-risk regime in Chapter III is separate and was deferred to 2 December 2027 for Annex III systems and 2 August 2028 for Annex I products.

When did Article 50 take effect?

On 2 August 2026. Machine-readable marking applies from that date to systems placed on the market on or after it, and from 2 December 2026 to systems already on the market (Art. 111(4)).

What are the penalties for an Article 50 breach?

Up to €15 million or 3% of total worldwide annual turnover, whichever is higher (Art. 99(4)). For SMEs, start-ups and small mid-caps the lower amount applies (Art. 99(6)-(6a)).

Does the EU AI Act apply if my company is not in the EU?

It can. The Act applies to AI systems placed on the EU market or whose output is used in the EU, so a non-EU company with EU customers can be in scope. Non-EU providers of high-risk systems and general-purpose AI models also need an EU authorised representative.

Is a visible "AI generated" label enough?

Not for the marking duty. Article 50(2) requires machine-readable markers embedded in the output. A visible label can help meet the separate deep-fake disclosure duty in Article 50(4), but it does not replace machine-readable marking.

Is hiring software automatically high-risk?

Recruitment and selection systems (Annex III 4(a)) and systems that support decisions about workers or monitor their performance (4(b)) are high-risk. A system that only performs a narrow procedural or preparatory task and does not profile people can fall outside high-risk under Art. 6(3), but the provider must document that assessment and register the system.

Did the Digital Omnibus delay the whole AI Act?

No. It deferred the high-risk obligations by sixteen months for Annex III systems and twelve months for Annex I products. Article 50 transparency, the Commission's GPAI enforcement powers and the fine tiers were not deferred, and it added two prohibitions that apply from 2 December 2026.