The European Commission proposal spans cybersecurity, data and privacy, while Parliament has barely begun discussions and the Council pushes for a general approach.
Critics warn haste could deepen fragmentation across 27 member states, especially on personal data definitions, cookie consent rules and the legal basis for AI training.
The overhaul aims to modernise EU tech rules and boost innovation, but policymakers are urged to prioritise legal certainty and consistent enforcement over quick simplification.
Is the EU's rush to simplify tech law a genuine boost for innovation or a dangerous rollback of digital privacy rights?
With its AI rules fast-tracked, is the EU creating a future where powerful tech outpaces fundamental data privacy protections?
EU AI Act Omnibus 2026: Political Agreement, Key Compromises, and Enforcement Powers
Overview
In May 2026, the European Commission, Parliament, and Council reached a political agreement on the AI Act Omnibus, resolving key conflicts between AI regulation and existing sectoral laws by limiting overlapping rules and exempting machinery regulation from direct AI Act requirements. The agreement introduced clear bans on AI systems creating harmful content like child sexual abuse material, extended regulatory relief to small mid-cap companies, and established tiered obligations for general-purpose AI. It also strengthened the AI Office with supervisory and enforcement powers, including fines up to 3% of global turnover. Supporting innovation, the Act offers SMEs simplified compliance, regulatory sandboxes, and financial aid, while aiming to set a global standard balancing innovation with fundamental rights protection.