The European Commission has published an annual report on its review of prohibitions on AI and high-risk AI systems

The European Commission has published the annual report required by Article 112(1) of the AI Act. This report focuses on reviewing the list of prohibited AI practices and high-risk use cases set out in Annex III.

The report outlines the applicable legal framework and methodology, as well as contributions received from Member States and stakeholders via the AI Board and public consultations.

The report identifies a potential regulatory gap concerning AI systems that generate non-consensual intimate content or child sexual abuse material. Currently, these systems are not covered by Article 5. However, the report also refers to the political agreement reached on 7 May 2026 between the Commission and the European Parliament as part of the simplified 'AI Omnibus' legislative package, which introduced a specific prohibition on these systems into the text of the Regulation.

The report notes that the provisions on prohibited practices came into force in February 2025, but the enforcement rules are not yet applicable. The assessment of whether to include other practices under Article 5 is still in the early stages, due to a lack of practical experience to date — specifically, the absence of reported breaches of this article of the Regulation.

Regarding high-risk systems, the report states that any evaluation of the effectiveness of the rules will depend on the publication of guidance on the classification of high-risk systems and the experience gained through its implementation. The regulatory sandboxes provided for in the AI Act are identified as important mechanisms for gathering future evidence.

The Commission indicates that certain AI systems will be monitored in subsequent reviews. These include self-help and emotional support chatbots, which the report identifies as areas where further evidence is needed.

Permanent link: https://digital-strategy.ec.europa.eu/en/policies/report-review-prohibitions-and-high-risk-ai

For more information on the European Artificial Intelligence Regulation and related initiatives, please refer to the Artificial Intelligence section of digitalGOV and the European Commission's dedicated Artificial Intelligence page.Parte superior do formulário