Primary obligations
Under the EU AI Act, every organisation is strictly prohibited from using certain harmful forms of AI-enabled manipulation or deception that meet specific legal criteria under Article 5. In addition, transparency rules are in place that require you to clearly communicate when someone is interacting with AI.
The EU’s AI Act legislation applies progressively, with a full roll-out of the main application milestones foreseen by 2 August 2028. The timeline takes into account the AI Act amendments introduced by Digital Omnibus on AI.
01 Aug 2024 - Entry into force
02 Feb 2025 - General provisions (definitions & AI literacy) and prohibitions apply
02 Aug 2025 - Rules for general-purpose AI apply and governance must be in place
02 Aug 2026 - The majority of rules of the AI Act come into force and enforcement starts for applicable rules
02 Dec 2026 - New prohibitions + Article 50(2) transition apply **
02 Aug 2027 - Member States should have at least one AI regulatory sandbox per country operational
02 Dec 2027 - Rules for high-risk AI systems in Annex III apply **
02 Aug 2028 - Rules for high-risk AI embedded in regulated products covered by Annex I apply **
** Following the Digital Omnibus on AI, certain provisions of the AI Act have been amended.
Achieving compliance
The AI Act Compliance Checker link below will walk you through a series of questions designed to assess your level of risk and outline your obligations under the EU AI Act. It includes useful explainers to help you complete it but if you need more information, please contact your relevant National Competent Authority from the list in the link below.
Related links