High-risk · Article 6 · Annex III
Guidelines, consultations, examples and sources for classifying high-risk AI systems.
Article 4a of the EU AI Act allows processing of special categories of personal data for bias detection and correction in AI systems, subject to strict safeguards for fundamental rights, particularly for high-risk AI systems.
Source: European Commission / AI Act Service Desk
According to the EU AI Act, providers and users of AI systems must implement measures to develop AI competence among their staff. AI competence includes understanding, critical evaluation, and responsible use of AI. The regulation does not prescribe standardized training or specific competence levels.
Source: Bundesnetzagentur
Annex III of the EU AI Act defines criteria for classifying AI systems as high-risk and sets detailed requirements for their development, deployment, and monitoring. It includes 83 specific articles covering risk management, data governance, transparency obligations, and conformity assessment procedures.
Source: European Commission
The document outlines the enforcement functions of the AI Office for GPAI models and AI systems. It clarifies responsibilities, investigative powers, sanctioning options, and monitoring tools. Enforcement begins on 2 August 2026.
Source: European Commission / European AI Office
On 19 June, the inaugural session of the AI Act Advisory Forum took place. The forum supports the implementation and enforcement of the AI Act, focusing on standardisation and classification of high-risk AI systems. It comprises 174 members from civil society, academia, and industry.
Source: European Commission
The Federal Network Agency prepares the implementation of the EU AI Act in Germany. The regulation governs AI systems based on risk, with strict requirements for transparency and safety at high risk levels. Effective from August 2027.
Source: Bundesnetzagentur
The European Commission provides draft guidelines on classifying high-risk AI systems under the AI Act. The document includes summaries and examples for AI systems in regulated products and sensitive areas.
Source: European Commission
The European Commission has appointed a Scientific Panel and an Advisory Forum to support enforcement of the AI Act. Both bodies advise the Commission's AI Office and national authorities on applying the rules. The Scientific Panel consists of 60 independent experts with experience in frontier AI, engineering, technical auditing, industry, and societal impact. The Advisory Forum provides independent technical expertise on issues such as standardisation and implementation challenges.
Source: European Commission
The EDPB and EDPS provide joint recommendations on data protection aspects of the European Biotech Act proposal. The proposal aims to strengthen the biotechnology industry while ensuring GDPR compliance in clinical trials and AI use.
Source: European Data Protection Board
The AI Act categorizes AI systems into four risk levels: unacceptable risk (banned), high risk (strictly regulated), limited risk (transparency obligations), and minimal risk (unregulated). High-risk systems require conformity assessments, while manipulative or surveillance systems are banned from 2025.
Source: Bundesnetzagentur
The EU AI Act promotes trustworthy AI, protects health and fundamental rights, and establishes uniform rules. It applies to providers, operators, and distributors of AI systems from August 2, 2025. Deadlines and guidelines are defined until 2028.
Source: Bundesnetzagentur
The European Commission is conducting a consultation until 23 July 2026 on draft guidelines for classifying high-risk AI systems. The goal is to assess the clarity, user-friendliness, and usefulness of the examples in the guidelines. The guidelines aim to support providers, deployers, and authorities in risk assessment.
Source: European Commission
The EU Commission has published draft guidelines for classifying high-risk AI systems and seeks stakeholder feedback. The guidelines aim to help providers and deployers assess whether a system qualifies as high-risk, complementing the AI Act requirements.
Source: European Commission
The European Commission has opened a consultation on draft guidelines for classifying high-risk AI systems. The goal is to assess the clarity and usefulness of the examples. The guidelines support providers, users, and market surveillance authorities in evaluating whether an AI system should be classified as high-risk. The deadline for feedback has been extended to 23 July 2026.
Source: European Commission
The EU Commission's draft guidelines assist providers, deployers and authorities in assessing whether an AI system qualifies as high-risk. They explain the criteria under Article 6 of the AI Act and provide practical examples.
Source: European Commission
The European Commission's draft guidelines assist providers and users of AI systems and market surveillance authorities in assessing whether an AI system should be classified as high-risk. They include practical examples and explain criteria under Article 6 of the EU AI Act.
Source: European Commission
The European Commission provides FAQs on implementing the EU AI Act. The regulation addresses risks to health, safety, and fundamental rights. It regulates prohibited AI systems, high-risk applications, transparency requirements, and general-purpose AI models.
Source: European Commission
The European Commission promotes harmonised standards for the EU AI Act to simplify compliance. These standards, developed by CEN, CENELEC, and ETSI, will be published from 2026 onwards. They provide legal certainty for providers of high-risk AI systems.
Source: European Commission
The European Commission promotes harmonized standards for high-risk AI systems to ensure legal certainty and set global benchmarks. CEN and CENELEC are developing standards in ten key areas, including risk management and transparency. The first harmonized standard was published on October 30, 2025.
Source: European Commission
The European Commission promotes harmonised standards for high-risk AI systems to ensure legal certainty and support innovation. These standards translate legal requirements into technical language, helping companies comply with the AI Act.
Source: European Commission
The European Commission provides FAQs on general-purpose AI models in the EU AI Act. It explains definitions, risks, obligations for providers, and transition periods. The focus is on transparency, copyright, and risk assessment.
Source: European Commission
The European Commission is preparing guidelines for the practical application of high-risk classification and transparency requirements under the AI Act to support compliance and provide clear instructions for businesses.
Source: European Commission / European AI Office
The EU AI Board is a central advisory body for implementing the AI Act. It coordinates national strategies, supports enforcement, and promotes ethical AI development. The Board consists of representatives from EU Member States and is supported by the AI Office.
Source: European Commission / European AI Office
The European Commission invites stakeholders to provide feedback on draft guidelines for transparency obligations of AI systems. The consultation runs from 8 May to 3 June 2026.
Source: European Commission / European AI Office
The European Commission has published guidelines on the definition of AI systems to facilitate the application of the first rules of the EU AI Act. The guidelines are non-binding and intended to evolve based on practical experiences.
Source: European Commission / DG CONNECT
The European Commission has published guidelines on prohibited AI practices under the AI Act. The guidelines clarify unacceptable practices such as harmful manipulation, social scoring, and real-time remote biometric identification. They aim to ensure consistent application of the AI Act across the EU.
Source: European Commission / DG CONNECT
The EU promotes research and industrial AI capacity while ensuring safety and fundamental rights. The AI Act establishes a four-tier risk framework. Initiatives like AI Factories and InvestAI support innovation and investments.
Source: European Commission / DG CONNECT
The EU AI Act is the first comprehensive legal framework for AI, addressing risks and positioning Europe as a global tech leader. It defines four risk levels, bans certain AI practices, and imposes strict requirements for high-risk systems. The Act comes into force gradually from 2025.
Source: European Commission / DG CONNECT
Article 7 of the EU AI Act empowers the European Commission to amend Annex III by adding, removing, or modifying high-risk AI system use cases. Changes are subject to specific criteria and conditions outlined in the regulation.
Source: European Commission / AI Act Service Desk
Article 6 of the EU AI Act defines criteria for classifying AI systems as high-risk. An AI system is considered high-risk if it is used as a safety component of a product requiring third-party conformity assessment or listed in Annex III. Exceptions apply under specific conditions.
Source: European Commission / AI Act Service Desk
Article 5 of the EU AI Act prohibits AI systems that violate fundamental rights, such as manipulative techniques, exploitation of vulnerabilities, social scoring, and real-time remote biometric identification. Two additional prohibitions apply from 2 December 2026.
Source: European Commission / AI Act Service Desk
Article 2 of the EU AI Act defines the scope of the regulation. It specifies who is subject to the regulation when offering, deploying, importing, or distributing AI systems or general-purpose AI models within the EU. Military applications, research activities, and natural persons using AI for personal purposes are excluded.
Source: European Commission / AI Act Service Desk
Article 1 of the EU AI Act defines the regulation's subject matter. It establishes the objective to improve the internal market and promote trustworthy AI while protecting fundamental rights and safety. The regulation sets harmonised rules for AI systems.
Source: European Commission / AI Act Service Desk
The FAQs address common questions on enforcement and governance of the EU AI Act. They cover the responsibilities of the AI Office, EDPS, and national authorities, the establishment of complaint and whistleblower tools, and the application of rules for GPAI models and high-risk AI systems.
Source: European Commission / AI Act Service Desk
The European Commission provides resources to support compliance with the EU AI Act, including fact sheets, guidelines, and webinars. The focus is on transparency obligations, codes of practice, and promoting AI literacy. Implementation begins on 2 August 2026.
Source: European Commission / AI Act Service Desk
The EU AI Act legislation is implemented progressively, with the full application of the main milestones by 2 August 2028. The timeline accounts for amendments introduced by the Digital Omnibus on AI.
Source: European Commission / AI Act Service Desk
The AI Act Explorer by the European Commission is a reference tool for the EU AI Act (Regulation (EU) 2024/1689). It enables searching for prohibited practices, risk classifications, transparency obligations, and requirements for AI systems. The document has been supplemented by the Digital Omnibus Regulation (EU 2026/1744).
Source: European Commission / AI Act Service Desk
Regulation (EU) 2024/1689 establishes harmonized rules for artificial intelligence to strengthen the internal market and ensure a high level of protection for health, safety, and fundamental rights. It regulates high-risk AI systems, protects personal data, and promotes trustworthy, human-centric AI.
Source: EUR-Lex / Europäische Union