EU Begins Enforcing AI Act With New Transparency Measures

The European Union has started enforcing the AI Act’s transparency and governance requirements, marking a new phase in the region’s approach to artificial intelligence oversight.

From 2 August 2026, the European Commission’s AI Office and national authorities began implementing the AI Act’s rules, including obligations for labelling content created by AI and informing users when they interact with AI systems. These measures are part of a broader regulatory framework designed to address the use and impact of artificial intelligence across the EU.

The enforcement responsibilities are distributed among the European Commission’s AI Office, national competent authorities designated by each Member State, and the European Data Protection Supervisor for AI systems used by EU institutions. This structure aims to coordinate oversight and ensure compliance with the Act’s provisions throughout the region.

The AI Act’s phased rollout began with its entry into force on 1 August 2024. General-purpose AI requirements started applying from 2 August 2025, followed by the main transparency and enforcement obligations from August 2026. Additional rules for high-risk AI systems are scheduled to come into effect in December 2027 and August 2028.

What the numbers show

• AI Act entered into force on 1 August 2024.
• Transparency and enforcement rules applied from 2 August 2026.
• Fines for violations can reach up to 7% of global annual turnover.

On 17 September 2026, the European AI Board convened under the Irish Presidency to discuss enforcement priorities for the AI Act and to review developments in international AI policy. This meeting formed part of ongoing efforts to coordinate regulatory approaches and share best practices among EU Member States.

International cooperation on AI governance has also been a focus for the EU. At the G7 summit in June 2026, leaders discussed ways to coordinate on advanced AI risks and considered a “trusted partners” scheme, which would allow non-U.S. countries access to advanced AI models. In July 2026, the EU delivered a statement at the UN Global Dialogue on AI Governance in Geneva, supporting a human-centric and rights-based approach to AI regulation.

The AI Act includes a range of penalties for non-compliance. According to the European Commission, fines can reach up to 7% of a company’s global annual turnover for violations involving banned AI practices. Other breaches may result in penalties of up to 3%, while providing incorrect information can incur fines up to 1.5% of global turnover.

The Act’s implementation timeline sets out several key milestones. Prohibitions and AI literacy measures took effect from 2 February 2025, governance and general-purpose AI obligations from 2 August 2025, transparency and enforcement from 2 August 2026, and high-risk AI rules are planned for December 2027 and August 2028.

* This article is based on publicly available information at the time of writing.