Right to explanation of an AI decision: what Article 86 of the AI Act gives you
If you are affected by a decision based (in part) on a high-risk AI system, Article 86 of the AI Act gives you the right to a clear explanation of the AI system's role and the main elements of the decision — from the deployer, on top of your GDPR rights.
Short answer: If an organisation takes a decision about you based (in part) on the output of a high-risk AI system listed in Annex III, and that decision has legal effects or similarly significantly affects you, Article 86 of the AI Act gives you the right to a clear and meaningful explanation: what role did the AI system play, and what were the main elements of the decision? You direct that right to the deployer, and it comes on top of your GDPR rights.
Who and when
The right applies to an affected person — the individual the decision is about. Three conditions together:
- the decision is taken by the deployer on the basis of the output of a high-risk AI system listed in Annex III (except the systems under point 2 of Annex III, critical infrastructure);
- the decision produces legal effects or affects the person in a similarly significant way;
- the person considers that this adversely affects their health, safety or fundamental rights.
Think of a rejected credit application, a refused insurance policy, a selection decision in recruitment, or the allocation of an essential service — in so far as a high-risk system was used.
What the explanation must contain
Not the full source code or model weights, but understandable information: the role of the AI system in the decision-making procedure and the main elements of the decision taken. The benchmark is intelligibility for a layperson, not technical completeness. "The computer said no" does not suffice; the organisation must be able to explain how the output shaped the decision.
Relationship to the GDPR
Article 86 applies "only to the extent that this right is not otherwise provided for under Union law" — it complements, it does not replace. GDPR Article 22 in principle prohibits a solely automated decision with significant effect and grants the right to human intervention and to make your view known. Article 86 of the AI Act is narrower in trigger (specifically high-risk Annex III), but explicit about the explanation itself. In practice they stack: a high-risk decision often engages both regimes at once — see also algorithmic decision-making in government.
Exceptions
The right is not absolute. Paragraph 2 provides that it does not apply where exceptions or restrictions follow from Union or national law (in compliance with Union law) — think of situations involving fraud detection, law enforcement or national security. And paragraph 3 makes it subsidiary: where an equivalent right to explanation already exists under other Union law, that applies. It is a floor, not an addition on top of existing rights.
What it requires of provider and deployer alike
The deployer can only explain what it understands. That depends on two other obligations: the provider must make clear, via the instructions for use (Art. 13), how the system works and is to be interpreted, and the system must keep logs (Art. 12) so that decisions are reconstructable. Without that basis, a meaningful explanation after the fact is impossible. So build the explanation process in before deployment, not when the first complaint arrives.
From when
Article 86 enters into application on 2 August 2026 (the general application date of the AI Act). Its practical effect runs in step with the high-risk obligations of Annex III — see the AI Act timeline of obligations for the exact phased dates.
What to do
- Map which decisions you base (in part) on high-risk Annex III systems, and who the affected persons are.
- Set up an explanation process: who provides the explanation on request, within what time, in plain language?
- Secure the basis from the provider: instructions for use (Art. 13) and logging (Art. 12) that make after-the-fact explanation possible.
- Align with your GDPR process: combine the Art. 86 request with the rights under GDPR Art. 22 and the information duties.
- Document the main elements of every relevant decision, so the explanation is reproducible and holds up under a complaint or review.
Article 86 makes explanation a right of the citizen, not a favour from the organisation. Whoever has their logging and instructions for use in order can comply; whoever defers it until the first request is left empty-handed.
Quellen
- https://eur-lex.europa.eu/eli/reg/2024/1689/oj
Regulation (EU) 2024/1689 (AI Act): Article 86 on the right to explanation of individual decision-making. - https://eur-lex.europa.eu/eli/reg/2016/679/oj
Regulation (EU) 2016/679 (GDPR): Article 22 on automated individual decision-making.
Lesen Sie auch
How the ECB supervises AI in eurozone banks: technology-neutral, existing frameworks, a generative-AI focus
For the 2026-2028 cycle the ECB places AI under its operational-resilience priority, and in February 2026 two Supervisory Board members set out the stance: with 85%+ of supervised banks using AI, govern it within existing frameworks rather than new rules, with a sharper focus on generative AI.
FRIA: Wann muss ich eine Grundrechte-Folgenabschätzung durchführen (KI-Verordnung Art. 27)?
Setzen Sie Hochrisiko-KI als öffentliche Stelle, als Anbieter öffentlicher Dienste oder für Kreditwürdigkeit oder Lebens- und Krankenversicherung ein? Dann verlangt Art. 27 KI-Verordnung vor der Inbetriebnahme eine Grundrechte-Folgenabschätzung (FRIA).
The Council of Europe AI Convention: the first binding AI treaty, ratified by the EU
The Council of Europe Framework Convention on AI is the first legally binding international AI treaty. The EU ratified it on 15 May 2026, after Parliament's consent on 11 March 2026. It binds states, not companies: principles and remedies that parties must transpose into national law.