Commission Article 112 AI Act review

Official metadata

Official title

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

Official reference

COM(2026) 234 final

Document type

Report

Document date

2026-05-20

Document date kind

Document issue

Document-date evidence

Publications Office of the European Union / EUR-Lex · EUR-Lex Document information > Date of document and Date of dispatch: 20/05/2026; COM(2026) 234 final Brussels masthead carries the same date. — Establishes retained document date 2026-05-20.

Publication date

2026-05-22

Publication-date evidence

European Commission, Directorate-General for Communications Networks, Content and Technology · Commission library Publication 22 May 2026, exact Article 112(1) report heading and exact download. — Establishes exact publication separately from issue, vote, adoption, signature, receipt, or consolidation.

Version label

Final

Legal status

Non binding

Language

en

Research classifications

These are researcher classifications used to organise and analyse the corpus; they are not official EU metadata.

Record level

Principal

Version status

Final

Temporal collection

Contemporary EU AI policy

Relevance class

Direct AI relevance

Policy stage

Implementation

Corpus tier

Core

Sectors

• General cross sector
• Health
• Employment and labour
• Migration asylum and border management
• Financial services
• Law enforcement
• Education
• Public administration
• Consumer protection
• Media and culture
• Critical infrastructure
• Cybersecurity

Policy membership

• AI Act implementation
• EU AI policy pathway

Classification evidence

• Relevance class: Direct AI relevance · Publications Office of the European Union / EUR-Lex · Official COM(2026) 234 final: the report applies Article 112(1) to Article 5 practices and Annex III uses spanning biometrics, infrastructure, education, employment, services, law enforcement and migration. — The cited provisions make AI the central subject.
• Sector tags: General cross sector · Publications Office of the European Union / EUR-Lex · Official COM(2026) 234 final: the report applies Article 112(1) to Article 5 practices and Annex III uses spanning biometrics, infrastructure, education, employment, services, law enforcement and migration. — The cited provisions substantiate the retained cross-sector classification.
• Provenance tags: Authored by an EU institution · Publications Office of the European Union / EUR-Lex · EUR-Lex Document information > Author: European Commission; COM(2026) 234 final masthead reads EUROPEAN COMMISSION. — Exact official metadata identifies the responsible EU institution or body.
• Provenance tags: Officially published · European Commission, Directorate-General for Communications Networks, Content and Technology · Commission library Publication 22 May 2026, exact Article 112(1) report heading and exact download. — Official EU page establishes publication of this exact manifestation.
• Sector tags: Education · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 36, printed p.8 / PDF p.9; paragraph 43, printed p.10 / PDF p.11. — Assesses personalised independent learning and discusses emotion recognition in educational institutions. The report does not recommend treating every personalised learning system as high-risk.
• Sector tags: Employment and labour · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 43, printed p.10 / PDF p.11; paragraph 58 footnote 36, printed p.16 / PDF p.17. — Discusses workplace emotion recognition and analyses reported recruitment, dismissal and employee-surveillance incidents. These are the report's descriptions and assessments, not independently verified incident findings.
• Sector tags: Migration asylum and border management · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 43, printed p.10 / PDF p.11; paragraph 58 footnote 36, printed p.16 / PDF p.17. — Examines calls to prohibit emotion recognition in migration contexts and describes migrant tracking/facial-identification incidents in border management. Further-monitoring conclusions are retained rather than converted to new prohibitions.
• Sector tags: Law enforcement · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 58 footnotes 36-37, printed p.16 / PDF p.17. — Analyses law-enforcement uses including domestic-violence victim risk assessment, suspect identification and traffic/public-event camera cases. Footnote 37 contains substantive comparative-risk reasoning, not merely a bibliographic reference.
• Sector tags: Health · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraphs 34-35, printed p.8 / PDF p.9; paragraph 46 second-page continuation, printed p.12 / PDF p.13. — Examines mental-health support chatbots, medical-device boundaries and continuing monitoring of potential harms. The report's conditional classifications are not presented as universal determinations.
• Sector tags: Public administration · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 37, printed pp.8-9 / PDF pp.9-10; paragraph 46, printed p.11 / PDF p.12. — Assesses public-authority authorisation decisions and discusses social-welfare/social-housing eligibility systems. Suggestions and monitoring commitments are distinguished from enacted amendments.
• Sector tags: Financial services · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 46, printed pp.11-12 / PDF pp.12-13. — Examines insurance categories beyond life/health insurance and AI-assisted debt collection by financial and other service providers. Financial-harm terminology alone is not the basis for the assignment.
• Sector tags: Consumer protection · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 43, printed p.10 / PDF p.11; paragraph 46, printed p.12 / PDF p.13; paragraph 57, printed p.15 / PDF p.16. — Assesses AI-enabled dark patterns, addictive design, debt-collection risks to vulnerable consumers and scams. Existing-law coverage and unresolved monitoring questions remain qualified.
• Sector tags: Cybersecurity · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 30 and footnote 8, printed p.7 / PDF p.8. — Assesses a proposed prohibition on AI intended to develop or distribute malware and explains the Commission's view on existing cybersecurity-law coverage and further monitoring.
• Sector tags: Critical infrastructure · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraph 30, printed p.7 / PDF p.8; paragraph 46 final bullet, printed p.12 / PDF p.13. — Directly discusses malware threats to critical infrastructure and proposals to extend coverage beyond safety components to reliability/continuity systems, with further monitoring rather than an adopted extension.
• Sector tags: Media and culture · Publications Office of the European Union / EUR-Lex · COM(2026) 234 final, paragraphs 57 and 59, printed pp.15-16 / PDF pp.16-17. — Analyses non-consensual synthetic intimate content and political disinformation using generated videos/images. This is substantive information-content risk analysis, not a claim of cultural-industry policy.

Production provenance

Provenance tags

• Authored by an EU institution
• Officially published

Named bibliographic authors

None recorded after review

Institutions and roles

• European Commission · Author · role evidence · EUR-Lex Document information > Author: European Commission; COM(2026) 234 final masthead reads EUROPEAN COMMISSION.

Official sources and identifiers

CELEX

52026DC0234

• Official source · Publisher/host: Publications Office of the European Union / EUR-Lex
• Official source · Publisher/host: European Commission, Directorate-General for Communications Networks, Content and Technology

Research assessment

Inclusion rationale

Verified official English text within the AI Act implementation and amendment corpus.

Researcher notes

Official metadata is transcribed from the cited source; corpus tier and policy-stage classification are researcher assessments.

Review status

verified

Relationships

No additional published relationships touch this record.

Verification

Expanded evidence review

Verified

Reviewed by

Yichen Hao

Evidence review date

2026-09-05T18:06:34Z

The evidence review date is not a separate human sign-off date. Review credit and evidence-review actions

• Last verified: 2026-09-04T00:00:00Z · EUR-Lex confirms COM(2026) 234 final, CELEX 52026DC0234, dated 20 May 2026, adopted pursuant to Article 112(1) of the AI Act.
• Last verified: 2026-09-05T18:06:34Z · Publication 22 May 2026; exact Article 112(1) report description and download.