European developments in the field of AI
The European Commission has launched a consultation on measuring the energy consumption, efficiency and emissions of AI models and systems, as part of a broader study supporting environmentally sustainable AI in the EU.
1.EU lawmakers have reached a political agreement on postponing the application of the AI Act under the Digital Omnibus AI, specifically in relation to:
• the obligation to label AI-generated content by technical means enabling its detection and traceability, until 2 December 2026;
• obligations for high-risk AI systems expressly listed in the AI Act, until 2 December 2027;
• obligations for AI systems used as safety components and subject to sectoral EU legislation on safety and market surveillance, until 2 August 2028.
2. The European Commission has published draft guidelines on the classification of high-risk AI systems under the AI Act, including practical examples. Determining whether a particular AI system falls within the category of a high-risk AI system is essential for establishing the scope of obligations under the AI Act. These guidelines are intended to clarify and harmonise the interpretation of Article 6 of the AI Act.
3. The European Parliament has reached an agreement on the wording of a ban on AI applications enabling the creation of sexualised images and videos of real persons without their consent. Following formal approval, this ban is expected to apply from 2 December 2026.
The European Commission has launched a consultation on measuring the energy consumption, efficiency and emissions of AI models and systems, as part of a broader study supporting environmentally sustainable AI in the EU. Providers of general-purpose AI models are already required to include known or estimated energy consumption in their technical documentation. The consultation, which is open until 1 June 2026, aims to provide a basis for developing criteria and a measurement framework for energy consumption in the field of AI.
