EU Countries to Coordinate Positions on AI Act Revisions by April 2026
Brussels (Brussels Morning Newspaper) January 12, 2026 – EU member states aim to finalize positions on proposed amendments to the AI Act by April 2026. The regulation, applicable since August 2024, is undergoing a review to address enforcement issues and technological developments. Coordination involves the Commission, Council, and Parliament, focusing on high-risk systems and general-purpose AI models.
AI Act Implementation Timeline and Phased Rollout
The Artificial Intelligence Act took effect on August 1, 2024, with prohibitions on practices like government social scoring applying immediately. Requirements for high-risk systems will begin in August 2026, while obligations for general-purpose AI started in August 2025. Codes of Practice for foundational models completed public consultation in late 2025.
National authorities are responsible for enforcing these rules, supported by the European AI Office. Member states are required to submit yearly reports on their progress. Stakeholder input from 2025 identified gaps in high-risk definitions and model evaluation criteria.
Key Amendment Proposals Under Member State Review
Discussions are centered on refining the high-risk system lists in Annex III, which cover biometric identification and education tools. Proposals are being considered to adjust thresholds for systemic risk in general-purpose models based on computing parameters. Enforcement updates may harmonize penalties, potentially reaching €35 million or 7% of turnover.
Council working parties under successive presidencies facilitate technical exchanges. France, Germany, and the Netherlands are contributing leading inputs on the balance between innovation and safety. Pilots from 2025 sandboxes will inform evidence-based changes.
Amendment Areas and Potential Revisions
- High-Risk AI: Fixed Annex III — Comitology updates
- GPAI Models: Risk assessments — Capability scaling
- SME Support: Basic exemptions — Raised thresholds
- Penalties: National variation — EU coordination
Privacy Enforcement and AI Rules
Privacy enforcement intersects with AI rules, as data protection authorities scrutinize training practices. Tim Green, an AI ethics commentator, noted in an X post that Meta has paused the rollout of its Meta AI in Europe following a request from the Irish Data Protection Commission due to privacy concerns over the use of Facebook and Instagram data for AI training. This represents a setback for EU AI innovation, yet it is critical for user consent standards.
Coordination Structures Among Member States
The AI Technical Working Party convenes regularly, with COREPER ambassadors assessing advances before April. The Commission provides impact assessments, and sandboxes operational in multiple states since 2025 are testing flexibilities. Italy and Spain have documented sandbox successes in voluntary compliance, while Germany is aligning proposals with GDPR frameworks. Over 200 organisations provided feedback during December 2025 dialogues.
The AI Pact counts 400 voluntary adherents advancing early adoption.
European Commission Oversight of Amendment Drive
The Commission exercises delegated authority for annex updates under scrutiny. Substantive changes follow the ordinary legislative procedure. The Executive Vice-President for tech sovereignty directs AI Office operations, with DG Connect drafting texts based on input from national experts. Codes for general-purpose AI will be finalized in the first quarter of 2026. The Act integrates with the Digital Services Act and Data Act provisions, with annual enforcement costs projected at €4-10 billion Union-wide.
Global Standards Influenced by EU Approach
EU rules are shaping international frameworks, with references in Japan and South Korea legislation. United States bills in 2026 are considering risk tiers, while the G7 Hiroshima commitments address generative risks. China prioritizes security in AI governance, and Singapore and Canada are implementing a tiered system modeled on the EU design. Member states are pursuing cross-border harmonization.
Consensus Challenges Facing April Deadline
Positions vary on expanding prohibitions and fine-tuning allowances. Smaller member states are seeking enforcement capacity aid, while major economies advocate flexibility. France is strengthening biometric safeguards, and Nordic countries are expanding research carve-outs. The March Competitiveness Council will review the status, coinciding with the Digital Decade assessment targeting 75% public AI adoption.
Stakeholder Input Shaping Revisions
BusinessEurope endorses burden reductions for startups. The Parliament’s internal market committee is tracking progress, while rights groups demand non-regression protections. National parliaments are evaluating subsidiarity, with the EESC delivering 2025 SME opinions. Sandboxes in 15 states are supplying deployment data.
