The landscape of AI regulation in Europe evolves. Several foundational rules—such as the ban on certain prohibited AI practices (Article 5) and the promotion of AI literacy (Article 4) – have been in effect since 2 February 2025.
The next significant milestone of the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) comes into force on 2 August 2025. This date marks the start of a critical compliance phase for organizations developing, deploying, or distributing AI systems within the EU.
Here are the main changes to be aware of:
- Notified Bodies & Conformity Assessment
Organizations will need to engage with designated notified bodies for the assessment of high-risk AI systems. These bodies will be appointed and supervised by notifying authorities. - General-Purpose AI Models (GPAI)
Providers of GPAI models must meet transparency obligations, provide technical documentation, and implement risk mitigation measures. - Governance & Oversight
National supervisory authorities will coordinate with the newly established European Artificial Intelligence Board (EAIB) to ensure consistent enforcement and guidance across the Union. - Enforcement & Penalties
Provisions on penalties for non-compliance take effect, excluding Articles 101 and 78 which will apply at a later stage. - Confidentiality Provisions
Rules on the handling of confidential information are outlined, though key provisions (e.g., Article 78) become binding at a later date.
Now is the time for organizations to review their AI governance strategies, establish clear documentation, and ensure alignment with the Act’s phased implementation schedule.