Checklist of what providers of high-risk AI systems must do under Article 16 AI Act: Chapter III requirements, quality management system, technical documentation, logs, conformity assessment, registration and post-market duties.
How high-risk AI systems are assessed under Article 43 AI Act - internal control (Annex VI) or notified body (Annex VII) - followed by the EU declaration of conformity, CE marking and registration in the EU database.
What providers of general-purpose AI (GPAI) models must do under Articles 53-55 AI Act: documentation, copyright policy, training-data summary, systemic-risk duties, and the Article 54 authorised representative for non-EU providers.
Article 50 AI Act transparency obligations for chatbots, machine-readable marking of AI-generated content, emotion recognition and deepfake disclosure - with the 2 August 2026 and 2 December 2026 dates.
The AI practices banned by Article 5 AI Act since 2 February 2025 - manipulation, social scoring, untargeted facial scraping, workplace emotion recognition and more - plus the non-consensual intimate imagery and CSAM bans added by Regulation (EU) 2026/1744.
What companies that use high-risk AI systems must do under Article 26 AI Act - instructions, human oversight, logs, informing workers and affected people - and who must carry out a fundamental rights impact assessment under Article 27.
When AI-enabled medical devices are high-risk under the AI Act, the 2 August 2028 date, a single conformity assessment through the MDR/IVDR notified body, and the separate MDR and AI Act authorised representative roles.
The AI Act fine levels under Articles 99 and 101: up to EUR 35 million or 7% for prohibited practices, EUR 15 million or 3% for other obligations including those of authorised representatives, and the SME rules.