Every AI Policy Ever · open in the interactive register
Published (non-binding, current)Non-binding · Report (government-commissioned)
Report from the Commission to the European Parliament and the Council: Report adopted pursuant to Article 112(1) of Regulation (EU) 2024/1689 on the need to review the list of prohibited AI practices and of high-risk AI systems listed in Annex III
Record EU-15 · COM(2026) 234 final
- Issued by
- European Commission
- Jurisdiction
- European Union
- Level · region
- Supranational · Europe
- Published / adopted
- 2026-05-20
- Language
- All 24 official EU languages
- Official identifier
- COM(2026) 234 final
Summary
The Commission's first report under Article 112(1) of the AI Act on whether the list of prohibited AI practices in Article 5 and the high-risk use cases in Annex III need to be amended. Drawing on consultation of Member States through the AI Board, a multi-stakeholder consultation and an analysis of AI incident databases, it identifies a potential regulatory gap for AI systems generating child sexual abuse material and non-consensual intimate content, which Article 5 does not prohibit. It flags other systems for monitoring, among them emotion recognition, AI companions and mental-health chatbots, and political disinformation.
Safety topics
prohibited practicesrisk-based classificationdeepfakes & synthetic mediachild safety
Status notes
Adopted on 20 May 2026 and addressed to the European Parliament and the Council. It does not review the list of areas in Annex III, which Article 112(2) requires by 2 August 2028. The report finds its assessment limited: when it was adopted, the rules on enforcing the prohibitions applied only from 2 August 2026, national competent authorities were still being designated and the Commission's guidelines on the classification of high-risk AI systems were still in preparation. It refers to the Commission's proposal for a Digital Omnibus on AI, which would tie the start of the high-risk rules to the availability of standards and support tools, no later than 2 December 2027 for Annex III systems. The next assessment under Article 112 is due in 2026. A summary of the report by Portugal's Agency for Technological Reform of the State says that the political agreement of 7 May 2026 between the Commission and the European Parliament on the AI Omnibus introduced a specific prohibition of AI systems generating non-consensual intimate content or child sexual abuse material into the AI Act.
Local copy
Archived. Last attempt 2026-10-17 18:18:31 UTC
Official source (third-party copy, not published) archive/EU-15/2026-10-11/official.html
Retrieved 2026-10-11 13:58:49 UTC
SHA-256 ed28717485c2ffa078c6cd7688bd24231f54456cb4dac5f9af8ad125e76e93c9
Verification
AThe instrument's own official text, or an official register, was stored and confirms the title, date and status. checked 2026-10-17 · official-url-fetched
Re-read in the recheck of 17 Oct 2026: the EUR-Lex HTML of COM(2026) 234 final (official, stored 11 Oct 2026) is the report's own full text; it confirms the title, the date (Brussels, 20.5.2026), the Article 112(1) basis, the methodology and the findings. The same EUR-Lex file is also stored as a source of EU-14 and EU-06. Stored in this run: a news item of 5 June 2026 by ARTE, the Portuguese Agency for Technological Reform of the State (digital.gov.pt), which summarises the report and reports the 7 May 2026 AI Omnibus agreement adding a prohibition of systems generating non-consensual intimate content or CSAM. That item says the report refers to the agreement, but the stored EUR-Lex text mentions the Digital Omnibus only as a proposal still under negotiation. Version 1's summary said the report concludes there is no immediate need to amend Article 5 or Annex III; the report uses that wording only for emotion recognition, and the summary now follows its conclusions. Version 1's date of 2 Dec 2026 for the new prohibition is not in this record's stored files and was removed.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-17: 7 field(s) supported.
History
- 2026-10-17notes: Adopted 20 May 2026. The gap it identified (CSAM and non-consensual intimate dee → Adopted on 20 May 2026 and addressed to the European Parliament and the Council. — rewritten from the report; the Digital Omnibus prohibition date (2 Dec 2026) removed; the report's own account of the Omnibus proposal added, and the 7 May 2026 agreement adding the prohibition from the Portuguese government agency's summary2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
- 2026-10-17related: [{"slug": "eu-ai-act", "id": "EU-03"}, {"slug": "eu-digital-omnibus-ai", "id": " → [{"id": "EU-03", "slug": "eu-ai-act"}, {"id": "EU-18", "slug": "eu-digital-omnib — EU-14 added: the report refers to the forthcoming high-risk classification guidelines2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
- 2026-10-17summary: The first annual assessment required by AI Act Art. 112(1). It concludes there i → The Commission's first report under Article 112(1) of the AI Act on whether the — rewritten to follow the report's conclusions: it notes a potential gap for CSAM and non-consensual intimate content and flags other systems for monitoring; 'no immediate need to amend' is the report's wording for emotion recognition only; the methodology added2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
- 2026-10-17title: Report from the Commission to the European Parliament and the Council on the nee → Report from the Commission to the European Parliament and the Council: Report ad — the report's own title from its cover page: version 1 left out 'Report adopted pursuant to Article 112(1) of Regulation (EU) 2024/1689'2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
Citation
European Commission. Report from the Commission to the European Parliament and the Council: Report adopted pursuant to Article 112(1) of Regulation (EU) 2024/1689 on the need to review the list of prohibited AI practices and of high-risk AI systems listed in Annex III [COM(2026) 234 final]. 2026-05-20. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX%3A52026DC0234. Status: Published (non-binding, current); verified 2026-10-17 (level A). Local copy archive/EU-15/2026-10-11/official.html retrieved 2026-10-11 13:58:49 UTC, SHA-256 ed28717485c2ffa078c6cd7688bd24231f54456cb4dac5f9af8ad125e76e93c9. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record EU-15.
Improve this record
Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.