EU AI Act: enforcement calendar

Europe’s Artificial Intelligence Act is no longer a distant rulebook: key obligations are already live, and the main enforcement date is August 2, 2026. (digital-strategy.ec.europa.eu) The law entered into force on August 1, 2024. Its first deadlines followed on February 2, 2025, when bans on prohibited AI practices and duties around AI literacy started to apply across the European Union. (digital-strategy.ec.europa.eu) Another major step landed on August 2, 2025, when obligations for providers of general-purpose AI models — the base models used to build chatbots, coding tools, and other systems — started to apply. The European Commission published guidelines on July 18, 2025, and a voluntary code of practice was published on July 10, 2025, to help companies meet those duties. (digital-strategy.ec.europa.eu, digital-strategy.ec.europa.eu) The big date now is August 2, 2026. That is when the Act becomes fully applicable in general, and the European Commission says the European Artificial Intelligence Office will supervise the most powerful general-purpose AI models while national authorities handle implementation inside member states. (digital-strategy.ec.europa.eu, digital-strategy.ec.europa.eu) The structure is split. The Artificial Intelligence Office sits inside the Commission and oversees general-purpose AI models, while national competent authorities enforce the Act domestically, creating a two-layer system for companies selling models and AI products in Europe. (digital-strategy.ec.europa.eu) That timetable overlaps with the General Data Protection Regulation, Europe’s privacy law. The International Association of Privacy Professionals said this week that the Artificial Intelligence Act works as a product-safety regime while the General Data Protection Regulation protects personal data and individual rights, leaving many companies to map both at once. (iapp.org) The overlap is not theoretical. The European Data Protection Supervisor said in November 2025 that joint guidance from the European Data Protection Board and the Commission on the General Data Protection Regulation and the Artificial Intelligence Act was being prepared, and the International Association of Privacy Professionals reported last month that cross-regulatory guidance was expected before the end of 2026. (iapp.org, iapp.org) There is still one later deadline after August 2026. The Commission’s materials say some provisions for high-risk AI systems and certain products continue phasing in until August 2, 2027, so compliance work does not end when next summer’s date arrives. (eur-lex.europa.eu, digital-strategy.ec.europa.eu) For providers, that means the calendar now runs in sequence: banned uses from February 2025, model-provider duties from August 2025, broad application and stronger supervision from August 2026, and final phase-ins into August 2027. Europe’s AI law has moved from drafting to deadlines. (digital-strategy.ec.europa.eu, eur-lex.europa.eu)