The Digital Markets Act (DMA), Regulation (EU) 2022/1925, marked its second anniversary of effective enforcement in March 2026. As the European Commission continues to expand investigations, issue specification decisions, and impose periodic penalty payments on designated gatekeepers, the policy debate has matured beyond celebration of the regulatory milestone toward a harder question: is the DMA actually delivering more contestable digital markets, or is it producing compliance theatre at the cost of European innovation?
The evidence to date is mixed — and that is precisely why the Commission's mandatory review under Article 53 DMA, due by 3 May 2026, deserves a serious, evidence-based reassessment rather than a reflexive expansion of scope.
What the DMA Promised — and Where It Stands
The DMA was designed to address structural concerns about the bargaining power of very large online platforms acting as intermediaries between business users and consumers. Following designation of the first gatekeepers in September 2023 — Alphabet, Amazon, Apple, ByteDance, Meta, and Microsoft, with Booking added in May 2024 — obligations under Articles 5, 6, and 7 became enforceable in March 2024.
By spring 2026, the Commission has opened formal non-compliance proceedings covering Apple's App Store anti-steering rules and choice-screen design, Alphabet's self-preferencing in Search and conduct in Google Play, and Meta's "pay-or-consent" advertising model. The April 2025 decisions fining Apple €500 million and Meta €200 million confirmed that the Commission is willing to use the DMA's enforcement teeth — penalties of up to 10% of global turnover, and 20% for repeat offenders.
The Compliance Burden — and Who Actually Pays
Less examined is the cumulative cost of compliance fragmentation. Apple has rolled out three distinct App Store regimes in the EU since March 2024, each followed by Commission objections. Google has rebuilt its Search results pages in Europe to comply with self-preferencing rules — and faced backlash from the very vertical comparison sites the DMA was meant to protect, several of whom report traffic losses after the redesign.
This is not an argument against enforcement. It is an argument for sequencing. As Mario Draghi observed in his September 2024 report on European competitiveness, the EU's overlapping regulatory stance towards tech can hamper innovation when duties under DMA, DSA, GDPR, the AI Act, and the Data Act are imposed without coordinated guidance. Enrico Letta's April 2024 Single Market report reached a similar conclusion: regulatory fragmentation is itself a barrier to European scale.
Innovation Costs Are Showing Up in the Data
Three signals deserve attention as the Commission prepares its review:
- Feature delays in the EU. Apple Intelligence, Google's most advanced Gemini features, and several Meta AI capabilities launched later in the EU than in other jurisdictions, with companies citing DMA interoperability and data-sharing uncertainty. Whatever one thinks of those rationales, the gap is measurable for European consumers.
- Investment trends. Bruegel's 2025 analysis of EU digital investment found that European venture funding in platform-adjacent sectors continued to lag the United States by roughly a factor of four, with gatekeeper-adjacent compliance cited by founders as a deterrent to building on EU-designated core platform services.
- Limited contestability gains. The Commission's own monitoring shows modest take-up of alternative app stores and sideloading on iOS, and limited switching away from designated search and messaging services. Headline market structure has not meaningfully shifted.
A Proportionate Path Forward
None of this argues for repeal. The DMA's core insight — that calibrated ex ante rules can be more efficient than decade-long ex post antitrust cases — remains sound. But a pro-innovation, proportionate enforcement posture should prioritise four corrections in the Article 53 review:
1. Sequence and consolidate guidance
The Commission should publish consolidated, cross-instrument guidance reconciling DMA, DSA, GDPR, and AI Act obligations. The April 2024 EDPB opinion on "consent or pay" models, the DMA preliminary findings on the same Meta conduct, and parallel national DPA proceedings have left gatekeepers and challengers alike uncertain about what compliance actually looks like.
2. Resist scope creep
Calls to designate additional services — cloud infrastructure, generative AI assistants, automotive operating systems — should be evaluated rigorously against the quantitative and qualitative thresholds in Article 3, not used as a vehicle to extend gatekeeper duties to emerging markets where contestability is still high. The Court of Justice's July 2024 judgment in ByteDance v Commission confirmed the Commission's discretion in designation, but discretion should be exercised with restraint where competitive dynamics are still forming.
3. Take innovation losses seriously in remedy design
Where compliance solutions visibly degrade product quality for European users — slower feature rollouts, weaker security guarantees, reduced integration — the Commission should treat that as a signal that the remedy is misspecified, not as an acceptable cost of enforcement. Article 8(2) DMA already contemplates dialogue on compliance measures; that dialogue should be substantive, not procedural.
4. Build in faster review of gatekeeper status
Three years is a long time in digital markets. The DMA permits removal of gatekeeper status under Article 4, but the procedural threshold is high. The review should consider whether services facing genuine new competitive entry — search confronted with generative AI alternatives, for example — should be reassessed on a faster cadence.
The Wider Stakes
Europe's regulatory ambition is not in question. Its execution is. The DMA was meant to make digital markets more contestable for European businesses and consumers. Two years into enforcement, the contestability gains are modest, the compliance costs are real, and the innovation gap with the United States and Asia continues to widen. A serious, evidence-based recalibration in 2026 is the difference between a landmark regulation and a cautionary tale.