From the journal

European Commission Publishes Draft Guidelines on High-Risk AI Classification, 2026

The European Commission has published draft guidelines on the classification of high-risk AI systems under Article 6 of Regulation (EU) 2024/1689 (the AI Act) and opened a targeted consultation for stakeholder feedback. The consultation deadline has been extended to 23 July 2026. The guidelines, not legally binding, will guide national market surveillance authorities and providers determining whether their AI systems fall into the high-risk category.

3 min read

The European Commission published draft guidelines on the classification of high-risk AI systems under the AI Act, Regulation (EU) 2024/1689, and opened a targeted consultation for stakeholder feedback. The guidelines are in draft form and not yet final: the Commission extended the consultation deadline to 23 July 2026, after which it will finalise and publish the guidelines, with final adoption expected by end of 2026. The guidelines are intended to support providers and deployers in self-classifying AI systems and to assist national market surveillance authorities in enforcement.

Article 6 of Regulation (EU) 2024/1689 sets the criteria for classifying an AI system as high-risk. The draft guidelines address two categories: first, AI systems intended as safety components of products covered by EU harmonisation legislation listed in Annex I that must undergo third-party conformity assessment; second, AI systems falling into one of the use-case areas listed in Annex III, which covers biometric identification, critical infrastructure, education, employment, essential services, law enforcement, migration, and administration of justice. The guidelines also address the Article 6(3) exclusion allowing providers to self-determine that a listed AI system does not pose a significant risk of harm.

AI system providers, deployers, importers, and distributors placing AI systems on the EU market must determine whether their systems qualify as high-risk, as that classification triggers mandatory conformity assessment, registration in the EU database under Article 71, and ongoing post-market monitoring obligations. The draft guidelines affect developers of systems in healthcare, HR, education technology, credit scoring, biometrics, border control, and critical infrastructure, all of which are Annex III areas. Providers that incorrectly classify a high-risk system as non-high-risk face enforcement action by national market surveillance authorities.

The draft guidelines note that not all AI systems listed under an Annex III area automatically qualify as high-risk: Article 6(3) permits providers to self-determine that their system does not pose a significant risk, provided specified conditions are met and the determination is documented in the technical file. The Commission has indicated that these guidelines will be supplemented by additional guidance on obligations applying to providers and deployers of confirmed high-risk systems. Consultation responses submitted by 23 July 2026 will be published and considered before finalisation.

Licentium advises AI system providers and deployers on AI Act compliance, including high-risk classification assessments and conformity procedure preparation. Organisations seeking to respond to the consultation or to assess their products against the draft guidelines are welcome to contact us. Work we undertake includes AI Act readiness assessments, high-risk classification analysis, conformity assessment preparation, technical documentation review, and EU AI regulatory strategy.

Source: European Commission, Draft Guidelines on the Classification of High-Risk AI Systems under the AI Act (targeted consultation open to 23 July 2026)

AI Regulatory

More from the journal

See all

Bank of England Publishes Draft Rules for Systemic Stablecoin Issuers

The Bank of England published a policy statement and draft Code of Practice setting out the regulatory regime for systemic stablecoin issuers in the UK in June 2026. The draft rules introduce a GBP 40 billion temporary issuance guardrail per stablecoin, permit issuers to hold up to 70% of reserves in short-term UK government debt, and require the balance in central bank deposits. The consultation closes 22 September 2026; the Bank intends to finalise the Code of Practice by end of 2026.

Connecticut Enacts AI Responsibility and Transparency Act, Effective October 2026

On 2 June 2026, Connecticut Governor Ned Lamont signed Senate Bill 5 into law as Public Act 26-15, the Connecticut Artificial Intelligence Responsibility and Transparency Act. The law creates disclosure obligations for employers using automated employment decision tools, governance requirements for frontier AI developers, product standards for AI companion systems, and safety obligations for online platforms serving minors. Most provisions take effect 1 October 2026; the Attorney General holds exclusive enforcement authority.

EU AI Act Article 50 Transparency Obligations Apply from 2 August 2026

Four categories of transparency obligations under Article 50 of Regulation (EU) 2024/1689 become binding on AI system providers and deployers across the EU on 2 August 2026. The European Commission published a final Code of Practice on Transparency of AI-Generated Content on 10 June 2026 as a voluntary compliance tool, alongside draft guidelines issued on 8 May 2026 that remain under finalisation. AI systems placed on the EU market before the August deadline have until 2 December 2026 to comply with the machine-readable marking obligation in Article 50(2).