Skip to main content
Praxikon
All answers

Direct answer

What does Article 87 of the AI Act say about reporting of infringements and protection of reporting persons?

Your question is about Article 87: reporting of infringements and protection of reporting persons. That obligation applies today. Whether your system actually falls under it depends on conditions you assess yourself.

You determine this yourself

  • The trigger is a report of an infringement of this Regulation, whatever the risk class of the system: a report about an AI system outside the high-risk category is covered just as much. The protection itself is not unconditional. It comes from Directive (EU) 2019/1937, which in Article 4 requires the person reporting to have obtained the information in a work-related context, and in Article 6(1)(a) requires reasonable grounds to believe that what was reported was true and fell within the scope of that Directive.
  • Article 87 creates no channel requirement. That requirement comes from Article 8 of Directive (EU) 2019/1937. Paragraph 1 places it on legal entities in the private and the public sector; paragraph 3 limits paragraph 1 in the private sector to entities with 50 or more workers. That threshold is not general, however. Paragraph 4 provides that the threshold in paragraph 3 shall not apply to entities falling within the scope of the Union acts referred to in Parts I.B and II of the Annex to that Directive, which cover financial services, anti-money laundering and transport safety among others. Paragraph 7 allows a Member State, following a risk assessment, to require entities with fewer than 50 workers as well. Paragraph 9 applies paragraph 1 to all legal entities in the public sector, with the option for a Member State to exempt municipalities under 10 000 inhabitants and other small public entities. Below fifty workers there is therefore not simply no channel requirement: it depends on the sector you fall in and on what your Member State has decided. The right to report and the protection of the person reporting exist in any event, through the external route of Article 10 of that Directive.
  • The material scope of Directive (EU) 2019/1937 runs through Article 2(1)(a), which refers to the Union acts listed in the Annex to that Directive. Regulation (EU) 2024/1689 was not added to that Annex: it makes the Directive applicable directly, in Article 87. National transposition law that ties its own scope to that same Annex, such as the Dutch Wet bescherming klokkenluiders, may therefore lag behind the Regulation. Whether a report about an AI system falls under national law as a result is not settled.

First step: Make sure a report about an AI system reaches your reporting channel.

Article 87 of the AI Act covers reporting of infringements and protection of reporting persons. The whistleblower Directive applies to the reporting of infringements of the AI Regulation and to the protection of the persons who report them. For an organisation that falls under that Directive this means a report about an AI system runs through the same protected channel as any other report, and a person reporting who meets the conditions of that Directive may not be disadvantaged for it. Whether you fall under that Directive is not stated in Article 87 but in national transposition law. This provision concerns the deployer and the provider of an AI system. This provision has applied since 2 August 2026.

The conclusion and your first steps

This applies now

Your first actions

  1. Make sure a report about an AI system reaches your reporting channel. Only for organisations that must already have a reporting arrangement. Make visible in it that an infringement of the AI Regulation is a reportable infringement, designate who receives such a report, agree how the identity of the person reporting stays out of the rest of the process, and record whether you handle anonymous reports.
Read the official sourceChecked on

General interpretation, not legal advice. Checked against Regulation (EU) 2024/1689 and the Digital Omnibus (EU) 2026/1744; the official source remains authoritative.

Full map for your situation

Source and locator

Every statement above rests on these texts. The locator points to the place in the text, the version and the date say which edition was checked.

Dataset praxikon:sys:registry:dataset:ai-act-implementation-graph 2.2.0, schema 1.5.0.

Execution

Embed this obligation in policy, process and ownership

A legal action only works when it is clear who performs it, where the evidence sits and when it is reviewed. Embed AI translates this obligation into working agreements, controls and ownership per system. The source interpretation above is separate from this commercial referral by Praxikon to an affiliated brand.

See the governance approach
Does this answer your question?
Zahed Ashkara, jurist and freelance AI & Privacy Consultant

Behind this page

Zahed Ashkara

Freelance AI & Privacy Consultant, jurist