The EU Artificial Intelligence (AI) Act was published in the Official Journal of the European Union on July 12, 2024, as “Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence.” Please see our earlier blog post for a general overview of the AI Act.
While the AI Act will generally apply starting on August 2, 2026, the exact milestones are quite nuanced and complex, with some provisions applying as early as February 2, 2025.
Below, we set out the key dates for the various operators, especially providers and deployers, as well as the dates by which the European Commission will have to prepare implementing acts, documentation and reports to assist the operators to ensure compliance with the AI Act.
----- 2024 -----
August 1, 2024 |
Entry into force of the AI Act (Article 113). This means that the AI Act becomes part of the EU legal order. It does not mean that the provisions of the AI Act are applicable. |
November 2, 2024 |
By this date, Member States must have identified the public authorities or bodies that supervise or enforce obligations under EU law protecting fundamental rights, including the right to nondiscrimination, in relation to the use of high-risk AI systems referred to in Annex III of the AI Act (Article 77(2)). |
----- 2025 -----
February 2, 2025 |
Chapters I and II will apply from this date (Article 113(a)). These include the general provisions (e.g., geographic scope, definitions) and the provisions on prohibited AI practices (see our client alert). Therefore, evaluating activities that might be considered “prohibited practices” under Article 5 of the AI Act should start as soon as possible. The general obligation to ensure a sufficient level of AI literacy of staff under Article 4 of the AI Act will also apply from this date. |
May 2, 2025 |
By this date, codes of practice for the implementation of general-purpose AI models and related obligations must be ready (Article 56(9)). These codes should support providers in achieving compliance with their duties relating to general-purpose AI models. |
August 2, 2025 |
From this date, Chapter III, Section 4 (Notifying authorities and notified bodies), Chapter V (General-purpose AI models), Chapter VII (Governance) and Chapter XII (Penalties) will apply (except for Article 101, which deals with fines for providers of general-purpose AI models).
|
----- 2026 -----
February 2, 2026 |
By this date, the European Commission will issue implementing acts creating a template for high-risk AI providers’ post-market monitoring plans, which should serve as the basis for said monitoring system established by Article 72. Similarly, the European Commission must, by this date, provide guidelines for the practical implementation of Article 6 concerning the classification of an AI system as high risk (Article 6(5)). |
August 2, 2026 |
This is the default date for the provisions of the AI Act being applicable. The obligations regarding high-risk AI systems will apply from this date, including risk and quality management systems, diligent data governance, technical documentation, recordkeeping, and transparency and information obligations. Chapter IV addresses operators of AI systems directly interacting with humans, generative AI systems, and emotion recognition or biometric categorization systems, introducing disclosure and information responsibilities (see our client alert). By this date, Member States must have implemented rules on penalties and other enforcement measures and notified the European Commission about them (Article 99).
At least one AI regulatory sandbox must be operational at a national level (Article 57(1)). |
----- 2027 -----
August 2, 2027 |
This is the ultimate deadline for AI systems covered by existing harmonization legislation (Article 113(c)) and for providers of general-purpose AI models that have been placed on the market up to 12 months after August 1, 2024, to comply with the AI Act. |
For more information on this or other AI-related matters, please contact one of the authors.