The EU Is Fighting Yesterday’s Tech War

The European Union has zero companies in the global top ten artificial intelligence developers. Zero. OpenAI, Google DeepMind, Anthropic, Meta AI, xAI — all American. Baidu, Alibaba, DeepSeek — all Chinese. Europe’s answer to this civilizational gap? More enforcement actions against Apple’s App Store.

That is the essence of Europe’s current tech crisis, and it is more damning than any single statistic. Brussels spent the better part of a decade building an extraordinary legal architecture to discipline the platform monopolies of the 2010s — the Google search dominance era, the Apple App Store stranglehold, the Meta social media empire — at the precise historical moment when a different, far more consequential technological revolution was gathering speed. The EU is enforcing rules on yesterday’s battlefield while losing tomorrow’s war. And the clock is running out.

How Brussels Built a Perfect Regulatory Machine for a Fight That Already Ended

The flagship instrument is the Digital Markets Act (DMA), which entered into force on November 1, 2022, and became fully applicable on May 2, 2023. It is, by most measures, the most ambitious attempt by any democratic government to regulate private technology companies in history. The European Commission designated six “gatekeepers” in September 2023 — Alphabet, Amazon, Apple, ByteDance, Meta, and Microsoft — subjecting their 22 core platform services to sweeping new obligations. Fines reach up to 10% of annual global turnover, escalating to 20% for repeat offenders, with potential forced structural breakup for systemic violators. Complementing it is the Digital Services Act (DSA), operational since February 17, 2024, covering content moderation for platforms with more than 45 million EU users.

The problem is not that these laws are wrong. It is that they are late — and pointed at the wrong target for the moment. Consider the competitive landscape the DMA was designed to address versus the one Europe actually faces in 2025:

The Mario Draghi Report on European Competitiveness, released September 9, 2024, delivered the verdict in language that should have shattered glass in Brussels: Europe needs an additional €800 billion per year in investment to keep pace with the US and China, and explicitly named artificial intelligence as the critical battleground. Draghi warned that “Europe is not investing enough in its digital and AI future and risks falling into a permanent low-growth trap.” That report was received with great fanfare. Then enforcement actions against Google and Apple continued as scheduled. For more on the geopolitical dimensions of technology competition, see our worldwide political news coverage.

The DMA’s Apple Settlement, Google’s Charges, and the AI Office Scramble Happening Right Now

The timeline of 2025 reveals an institution in genuine tension with itself — simultaneously doubling down on yesterday’s regulation while scrambling to appear relevant to tomorrow’s challenges.

Here is what has actually happened since January 2025:

• March 2025: The European Commission reached a preliminary DMA settlement with Apple over App Store violations, requiring Apple to allow third-party app marketplaces — a concession critics immediately labeled minimal and years overdue, given that US courts had already moved on the same issue.
• March 2024 (enforcement continuing through 2025): Google faced its first formal DMA non-compliance charge related to search self-preferencing — a complaint that antitrust economists were raising about Google Search back in 2010.
• February 2025: The EU AI Office opened its first operations, tasked with overseeing General Purpose AI (GPAI) models under the AI Act — but staffed at a fraction of the level needed, with a mandate that vastly outstrips its institutional capacity.
• Early 2025: The Commission launched a DMA “simplification” review under sustained pressure from France and Germany, signaling that member states now openly question whether the regulatory burden is strangling European AI investment at exactly the wrong moment.
• Q1 2025: US Trade Representative Jamieson Greer formally warned that DMA and DSA enforcement targeting American tech companies could trigger retaliatory trade measures — framing European digital regulation explicitly as economic protectionism. This is the same geopolitical dynamic playing out across a broader pattern where political institutions are increasingly perceived as threats to economic security.
• August 2026 (approaching fast): The AI Act’s GPAI provisions — covering foundational models like GPT-5 and Gemini Ultra — phase in fully, representing the next major regulatory flashpoint and potentially the most consequential compliance fight in European tech history.

The EU’s largest homegrown AI investment, Mistral AI of France, was valued at approximately €6 billion after a June 2024 funding round. That sounds impressive until you note that OpenAI hit a $157 billion valuation in October 2024. Europe’s flagship AI champion is roughly 4% the size of one American competitor. The math does not lie.

Vestager’s Legacy, Ribera’s Inheritance, and Von der Leyen’s Impossible Pivot: The Architects of Europe’s Dilemma

Three figures define how Europe arrived here, and where it might go.

Margrethe Vestager

Margrethe Vestager, EU Executive Vice President for A Europe Fit for the Digital Age from 2019 to 2024, is the intellectual architect of everything Brussels has done in tech regulation for the past decade. She levied over €8 billion in competition fines against Google alone between 2017 and 2019, forcing interoperability requirements, blocking mergers, and establishing the conceptual framework that became the DMA. Her work was genuinely significant. Google did abuse its dominance. Apple’s App Store margins were genuinely predatory. But Vestager was solving for the technology industry as it existed in 2015, not the one taking shape in 2023. By the time the DMA reached full applicability, the existential competitive threat had fundamentally changed shape. History may judge her as someone who won the battle brilliantly while losing the war by misdirection.

Teresa Ribera

Teresa Ribera, EU Competition Commissioner since December 2024, has inherited a regulatory machine of enormous power and ambiguous purpose. She must simultaneously enforce the DMA and DSA — both politically necessary and legally mandated — while navigating mounting pressure from industry and member states to create innovation-friendly carve-outs for European AI development. That is not a contradiction she can resolve with a single speech. It requires a fundamental strategic reorientation that Brussels has not yet made.

Ursula von der Leyen

Ursula von der Leyen, European Commission President, has recently pivoted hard toward the language of “competitiveness” — commissioning the Draghi Report, speaking about European AI champions, and signaling that the regulatory posture may need adjustment. The pivot is real but insufficient. Von der Leyen is trying to maintain the political credibility of the DMA/DSA framework while simultaneously arguing that Europe needs to build what those laws, in their current application, may be slowing down. Saying you want both regulation and competitiveness without explaining which takes priority when they conflict is not a strategy. It is a press release.

Why Both the Regulators and the Tech Apologists Are Misleading You About Europe’s AI Crisis

The debate about EU tech regulation has calcified into two deeply unsatisfying camps, and both are dodging the hardest questions.

The pro-regulation camp is not wrong about its diagnosis of historical abuses. Cambridge Analytica was real. Google’s self-preferencing cost European competitors genuine market share. Apple’s 30% App Store cut was extractive rent-seeking dressed up as platform governance. These were legitimate harms requiring legitimate regulatory responses. The problem is that the pro-regulation camp refuses to acknowledge the timing and opportunity cost. Every compliance lawyer hired by a European AI startup to navigate DMA obligations is a machine learning engineer not hired. Every regulatory uncertainty that delays a European venture capital round is a US competitor that raises another $500 million instead.

The competitiveness reform camp — now growing rapidly, led by France and Germany — makes the opposite error. It frames European regulation as the primary cause of Europe’s AI gap, which flatters European policymakers because it implies the solution is simply to regulate less. It is not that simple. The US AI ecosystem did not produce OpenAI, Anthropic, and Google DeepMind because America has fewer rules. It produced them because of decades of DARPA investment in foundational research, Stanford and MIT creating genuine research-to-commercialization pipelines, and a venture capital ecosystem willing to absorb catastrophic losses at scale. Europe has none of those structural advantages, and removing the DMA does not create them.

The American tech industry’s preferred framing — that DMA and DSA are protectionism by another name — is the most cynically self-serving of the three positions. Companies like Google, Apple, and Meta have spent hundreds of millions lobbying against DMA requirements while simultaneously enjoying the benefits of European market access. Their argument that regulation harms innovation would be more persuasive if they had not spent fifteen years using the very monopoly power the DMA targets to crush European competitors.

What nobody wants to say plainly: Europe is losing the AI race for structural reasons that predate the DMA, that regulation has made modestly worse at the margin, and that no single policy adjustment will fix. The honest answer requires an €800 billion per year investment program, a genuine European research ecosystem, and a willingness to let homegrown companies scale before regulating them — all three simultaneously. That is hard. Blaming the DMA is easier.

Four Scenarios for How the EU’s Tech War Miscalculation Resolves by 2028

The trajectory is not fixed. Four distinct scenarios are plausible between now and 2028, ranging from managed adaptation to full strategic collapse.

• Scenario 1 — Regulatory Pivot Succeeds: The DMA “simplification” review produces meaningful exemptions for European AI startups. The EU AI Office develops genuine technical competence. France’s “European AI Champions” industrial strategy attracts sufficient capital to grow Mistral AI and its peers to competitive scale. Probability: Low-to-moderate. Requires political will that has not yet materialized at sufficient scale.
• Scenario 2 — Enforcement Collision with Washington: DMA and DSA enforcement actions against US companies escalate through 2025-2026. The Trump administration converts Greer’s warnings into actual trade measures — tariffs or market access restrictions — tying EU tech regulation directly to the broader transatlantic trade war. European companies caught in the crossfire face the worst of both worlds: American retaliation and continued regulatory compliance costs at home.
• Scenario 3 — GPAI Flashpoint in August 2026: The AI Act’s General Purpose AI provisions trigger a compliance crisis when they hit foundational models fully. OpenAI, Google, and Anthropic face expensive transparency and auditing requirements that slow European deployment of their most advanced models. European businesses, already behind on AI adoption, fall further behind. The EU is blamed — fairly — for accelerating its own AI dependency on US platforms.
• Scenario 4 — Structural Dependency Accepted: The EU quietly accepts permanent dependency on American hyperscalers — AWS, Azure, Google Cloud — for AI infrastructure, while maintaining the DMA and DSA as consumer protection mechanisms rather than tools of strategic competition. Europe becomes the world’s most sophisticated regulatory zone for technologies it does not build. This is the most likely scenario, and nobody in Brussels will say so out loud.

The Draghi Report forced the debate onto the main agenda. That matters. But a debate forced onto the agenda in September 2024 that has produced a DMA simplification review and an understaffed AI Office by mid-2025 is not a debate that is winning. Europe has the legal infrastructure to police the tech giants of 2015 with extraordinary precision. What it does not have — and what no regulatory reform package will automatically create — is the foundational research investment, the patient capital, and the political tolerance for failure that producing globally competitive AI requires. Brussels is still polishing its rulebook for a game the rest of the world has already moved past.

The question Europe must answer — honestly, urgently, without the usual political hedging — is whether it wants to be the world’s most rigorous technology regulator or a genuine technology power. Because the evidence of the last five years suggests it cannot be both at the same time, and the window for choosing is closing faster than any Commission working group wants to admit.

About The Author

Sophie Gesture

Political writer and analyst focused on global affairs, public policy, and emerging political trends.

See author's posts