The European Parliament debated the Digital Omnibus on AI on 15 June 2026, with MEPs diverging on how to simplify harmonised AI rules. EPP's Axel Voss argued for lighter compliance burdens on SMEs, warning that overregulation could stifle innovation. S&D rapporteur Brando Benifei pushed back, insisting that fundamental rights protections must not be weakened. Renew Europe's Dragoș Tudorache aligned with Voss on reducing red tape but stressed the need for clear liability rules. Greens-EFA MEP Alexandra Geese questioned whether the simplification proposals would undermine the AI Act's risk-based approach. The Left's Cornelia Ernst called for stronger enforcement mechanisms, particularly for high-risk AI systems.
Executive Vice-President for Tech Sovereignty Henna Virkkunen defended the Commission's proposal as a balanced approach to cut administrative costs without lowering safety standards. Divergences centred on the scope of exemptions for general-purpose AI and the timeline for conformity assessments. The file now moves to committee-level negotiations, with a plenary vote expected in autumn.
Stakeholder impact: AI developers and SMEs would benefit from reduced compliance costs under Voss's approach, but consumer groups and national supervisory authorities could face weaker oversight if exemptions are broadened. Benifei's position would preserve stronger rights protections, potentially increasing administrative burdens for high-risk AI deployers. The final outcome will shape the balance between innovation and fundamental rights in the EU's AI governance.