Home / EU AI News / EU AI & Sovereignty — June 28, 2026

EU AI & Sovereignty — June 28, 2026

EU AI & Sovereignty — June 28, 2026

EU AI & Sovereignty News — June 28, 2026

This week marked a watershed moment for European AI governance and digital sovereignty. With the European Parliament’s approval of the Digital Omnibus on June 16 and the Commission’s landmark Tech Sovereignty Package on June 3, the EU reshaped its AI Act and unveiled a long-term vision for technological autonomy in a single month. Here are the top 10 developments.

  1. European Parliament Approves Digital Omnibus on AI (June 16)
    The European Parliament voted 423-57 (174 abstentions) in favour of the Digital Omnibus — the first targeted amendments to the AI Act. The approval follows the trilogue agreement of May 7, 2026. The Omnibus introduces simplification measures, delays for high-risk compliance, and a ban on nudifier apps.
    Context: This is the first major revision of the AI Act since its adoption, reflecting the tension between fostering innovation and maintaining robust safeguards.
    Source: europarl.europa.eu — EP press room

  2. EU Tech Sovereignty Package Unveiled (June 3)
    The European Commission launched the European Technological Sovereignty Package with four pillars: Chips Act 2.0, Cloud & AI Development Act, a new EU Open Source Strategy, and a strategic roadmap for digital infrastructure. Together they form the EU’s most ambitious push for digital autonomy to date.
    Context: The package explicitly targets dependency on non-European technologies in critical sectors, from semiconductors to cloud computing.
    Source: ec.europa.eu — Europese Commissie

  3. AI Act Delay for High-Risk Systems Confirmed
    The AI Act is amended to give businesses more time to comply with high-risk AI obligations. The original deadline of August 2, 2026 is pushed back with a phased implementation. Critics argue the delay represents a watering down of rules and claim ‘Europe is caving to Big Tech.’
    Context: This is one of the most contentious elements of the Digital Omnibus — industry groups lobbied heavily for more time while civil society organisations warn of regulatory capture.
    Source: Reuters / Morgan Lewis

  4. Nudifier Apps and CSAM Production Banned (Effective December 2, 2026)
    From December 2, 2026, offering, selling, or using AI systems that generate realistic intimate images without consent becomes punishable. AI-generated child sexual abuse material (CSAM) is also prohibited. This is one of the most consequential new prohibitions under the amended AI Act.
    Context: The nudifier ban addresses the rapid proliferation of deepfake and image-generating AI tools that have caused widespread harm, particularly to women and minors.
    Source: Ogletree / Global Policy Watch

  5. Cloud & AI Development Act (CAIDA) Proposed
    Part of the Tech Sovereignty Package: CAIDA aims to protect sensitive data (health, finance, justice) of EU governments from US parent companies. This is one of the most far-reaching proposals for digital sovereignty and could fundamentally reshape how European public institutions procure cloud and AI services.
    Context: CAIDA directly targets the current dominance of US hyperscalers (AWS, Azure, GCP) in European public-sector cloud contracts.
    Source: Innobu / CNBC

  6. Chips Act 2.0: EU Strengthens Semiconductor Base for AI Ambitions
    Chips Act 2.0 focuses on advanced semiconductor production in Europe, with a specific emphasis on AI chips. The goal is to reduce dependency on Asia and the US and strengthen the AI supply chain within Europe.
    Context: Following the original Chips Act’s modest success, version 2.0 allocates additional funding and targets leading-edge nodes needed for next-generation AI hardware.
    Source: AENEAS Office / EC

  7. EU Open Source Strategy Published as Part of Sovereignty Package
    The Commission published COM(2026) 503 — the first EU-wide Open Source Strategy, designed to strengthen European open-source AI ecosystems, sovereign clouds, and digital infrastructure as a strategic pillar of technological autonomy.
    Context: The strategy acknowledges that open source is not just a cost-saving measure but a sovereignty instrument, reducing lock-in to proprietary platforms.
    Source: EC digital-strategy

  8. CADA: New Proposal for Digital Sovereignty Rules
    The CADA proposal (Competitiveness and Digital Autonomy Act) provides legal instruments for the EU and member states to commit to actions that improve competitiveness and strengthen digital sovereignty. Recently published and in consultation.
    Context: CADA complements CAIDA — while CAIDA focuses on public-sector cloud procurement, CADA takes a broader view covering digital infrastructure and industrial competitiveness.
    Source: NautaDutilh

  9. AI Act Simplification: Less Overlap, More Innovation-Friendly
    The AI Omnibus streamlines the AI Act by reducing overlap with existing regulations (including the Machinery Directive). The goal is to lower the administrative burden and stimulate innovation while maintaining the risk-based approach.
    Context: Businesses had complained bitterly about regulatory duplication — a single AI system could fall under both the AI Act and sector-specific legislation — and the Omnibus addresses this.
    Source: EP Think Tank / Stibbe

  10. AI Act Transparency Rules Take Effect (August 2026)
    From August 2026, the AI Act’s transparency obligations come into force: companies must disclose when AI systems are used in consumer interactions, and AI-generated content must be labelled. The prohibitions on unacceptable-risk AI also go live.
    Context: These represent the first enforceable AI Act provisions to actually apply to businesses, setting a compliance precedent for the rules that follow.
    Source: EC Shaping Europe’s Digital Future

Key Trends This Month

  • Acceleration vs. caution: The EU pushes forward on sovereignty (Tech Sovereignty Package, CAIDA, Chips Act 2.0) while simultaneously extending compliance timelines for high-risk AI — a dual-speed approach that defines the current moment.
  • Open source as strategy: The EU’s first Open Source Strategy elevates open-source AI from a community practice to a geostrategic instrument, with implications for sovereign cloud and AI infrastructure funding.
  • US-EU tech decoupling accelerates: CAIDA’s explicit targeting of US parent companies for sensitive public-sector data signals a deepening of the Atlantic tech divide.
  • Nudifier ban as precedent: The prohibition on AI-generated non-consensual intimate images sets a global benchmark, likely influencing legislation in the UK, Canada, and Australia.
  • Administrative simplification: The Omnibus reduces regulatory overlap, suggesting the EU heard industry concerns and is willing to adjust — but critics say the core risk-based architecture remains unwieldy.
  • Semiconductor sovereignty: Chips Act 2.0 targets leading-edge AI chips specifically, acknowledging that Europe cannot achieve AI sovereignty without its own advanced fab capacity.

This overview was compiled on June 28, 2026 by the Hermes AI news assistant. Sources accessed: europarl.europa.eu, ec.europa.eu, Reuters, Ogletree, Innobu, CNBC, AENEAS, NautaDutilh, EP Think Tank, and Stibbe.

Wat vond je van dit bericht?

Leave a Reply

Your email address will not be published. Required fields are marked *