Every few weeks someone publishes the same article. Europe is losing the AI race. Europe missed the boat. Europe regulated itself into irrelevance while America and China built the future. The numbers get cited, the graphs get posted, and everyone nods.
The numbers are real. I am not going to pretend otherwise.
But there is a question hiding under all of it that almost nobody asks. Behind in what race? A race has a finish line. And the people declaring Europe the loser have quietly decided where the finish line is without telling you they decided.
I am not here to defend Brussels. I run infrastructure for a living, and I have felt the dead weight of compliance that exists to protect a process instead of a person. I am also not here to cheer for Silicon Valley. Both of them have blind spots, and both of them are selling you a story.
Here is the story I want to tell instead. The AI race has two scoreboards. America is winning the one it chose. Europe is keeping a different one. And the most dangerous word in the whole conversation is “innovation,” because of how often it gets used to dodge a much older question: all this power, in service of what.
1. The Scoreboard Everyone Points At#
Start with the strongest version of the “Europe is behind” case, because it is strong.
In 2024, private AI investment in the United States hit $109.1 billion. That is nearly twelve times China’s $9.3 billion and twenty-four times the United Kingdom’s $4.5 billion (Stanford HAI, 2025). American institutions released 40 notable AI models that year. China released 15. Europe released three (Stanford HAI, 2025).
The structural picture is worse than one year of data. Mario Draghi, the former head of the European Central Bank, wrote the EU’s own competitiveness post-mortem in 2024, and his line is brutal. “There is no EU company with a market capitalisation over EUR 100 billion that has been set up from scratch in the last fifty years” (Draghi, 2024). All six American companies worth more than a trillion euros were built in that same window. Only four of the world’s top fifty tech companies are European, and EU firms spent roughly 270 billion euros less on research and development than American firms in 2021 alone (Draghi, 2024).
Look at the champions. Europe’s most valuable company is now ASML, the Dutch firm that makes the lithography machines every advanced chip on earth passes through, worth around $670 billion (James, 2026). That is a genuine triumph. It is also still smaller than several American chip firms, and a rounding error next to the more than $22 trillion of market value sitting in America’s seven largest tech companies (The Motley Fool, 2026). DeepMind, the best AI lab Europe ever produced, sits in London and has belonged to Google since 2014 (Google DeepMind, n.d.). Mistral, France’s great hope, raised 1.7 billion euros in 2025 at a valuation near $14 billion, and the round was led by ASML (Browne, 2025). Europe’s flagship AI startup had to be anchored by Europe’s flagship chip company, and it is still worth a sliver of a single American lab.
So yes. On this scoreboard, Europe is behind. Not unlucky. Behind. Anyone who tells you otherwise is not looking at the board.
2. Behind in What?#
Now ask the question the graphs are designed to skip. What is the race for?
America has been refreshingly honest about this. In July 2025 the White House released “America’s AI Action Plan.” It opens by calling the moment “a race to achieve global dominance in artificial intelligence,” and its very first policy heading is “Remove Red Tape and Onerous Regulation.” The private sector, it says, “must be unencumbered by bureaucratic red tape” (White House, 2025). Six months earlier the administration had already torn up the previous executive order on AI and replaced it with one titled “Removing Barriers to American Leadership in Artificial Intelligence” (Trump, 2025).
Vice President JD Vance carried the message to Europe in person. Standing in Paris in February 2025, at a summit the prior administration had built around AI safety, he reframed it in his opening breath. “I’m not here this morning to talk about AI safety,” he said. “I’m here to talk about AI opportunity.” He warned that “excessive regulation of the AI sector could kill a transformative industry just as it’s taking off.” And he compressed the whole thesis into one line. “The AI future is not going to be won by hand-wringing about safety. It will be won by building” (Vance, 2025). The United States and the United Kingdom then declined to sign the summit’s final statement, which committed its signatories to “inclusive and sustainable” and “human-centric” AI (AI Action Summit, 2025).
That is a coherent goal. Dominance. Growth. Speed. Build the biggest thing fastest and let the rest sort itself out. I have real sympathy for parts of it. But notice what it is. It is a choice of finish line. It is not the only one. And a scoreboard that counts dollars invested and models shipped is perfectly designed to make the people optimizing for dollars and models look like the only serious players in the room.
3. Europe Is Keeping a Different Score#
Europe is not failing to win the American race. Europe is running toward a different line, and it wrote the line down.
The EU AI Act, the world’s first comprehensive AI law, states its own purpose in Article 1. It exists to promote “human-centric and trustworthy” AI “while ensuring a high level of protection of health, safety, fundamental rights enshrined in the Charter” (European Union, 2024). Read that again. The success criterion written into the statute is not market share. It is the protection of human beings.
That Charter is not decoration. Its first article reads, in full, “Human dignity is inviolable. It must be respected and protected.” Its eighth article makes the protection of your personal data a fundamental right in its own name (Charter of Fundamental Rights, 2000). This is why Europe guards your data the way America guards free speech. It is a starting axiom, not a feature request. It is also why the EU was willing to fine Meta 1.2 billion euros in a single ruling for shipping European data to American servers improperly (European Data Protection Board, 2023).
The people who built the framework say the quiet part out loud. “Trust is a must, not a nice to have,” said Margrethe Vestager when the Act was proposed. Her colleague Thierry Breton put it even more plainly. “AI is a means, not an end” (European Commission, 2021).
That sentence is the whole argument. Hold onto it.
And this is not some fringe Brussels obsession. The thirty-eight democracies of the OECD, the United States among them, agreed back in 2019 that AI should respect “human-centred values and fairness,” including “human dignity” and “privacy” (OECD, 2024). Everybody signed the principle. The fight is over who still acts like they meant it.
4. Growth for Growth’s Sake#
Here are the first principles everyone keeps stepping over in the rush to ship.
Innovation is a means. Investment is a means. Compute, models, market cap, all of it, means. The writer Edward Abbey gave us the warning a generation ago. “Growth for the sake of growth is the ideology of the cancer cell” (Abbey, 1977). A tumor is, after all, extremely innovative. It scales. It disrupts. It optimizes for one metric with total commitment and no regard for the body it is killing.
A means has to serve an end. So what is the end? The oldest answer in the Western canon is still the best one. Aristotle called it eudaimonia, usually translated as flourishing. It is the one thing we want purely for its own sake, the thing for whose sake we want every other thing (Kraut, 2022). Money is for flourishing. Technology is for flourishing. A nation’s entire economy is, in the final accounting, for the flourishing of the actual people living inside it.
Now hold the two scoreboards up against that standard.
We measure dollars invested. We measure models shipped. We measure compute and parameters and valuations. We do not, on the dominance scoreboard, measure whether a single human being is better off. We have built an exquisite instrument for tracking the means, and we have thrown away the ruler for the end.
Here is the part the booster crowd hates. Even America’s own deregulatory executive order says the point of AI dominance is “to promote human flourishing” (Trump, 2025). Good. Everyone agrees on the end. The argument is only about whether you are allowed to say the end out loud while the building is going on, or whether saying it makes you a hand-wringer who is bad at the race.
5. The Cost of Forgetting#
Forgetting first principles has a price, and you can already read it on the American ledger.
The people are not comfortable. A Pew survey found that 52 percent of Americans are more concerned than excited about AI, up from 38 percent two years earlier, and 53 percent say AI does more to hurt their privacy than to help it (Tyson & Kikuchi, 2023). Stanford’s own index found that only 39 percent of Americans see AI products as more beneficial than harmful, against 83 percent in China (Stanford HAI, 2025). The public is not asking for dominance. They are asking whether they can trust this thing with their kids and their data. That is a first-principles question, and “move fast” is not an answer to it.
Even America’s institutions flinched. When Congress tried to slip a ten-year ban on all state AI regulation into the 2025 budget bill, the Senate stripped it out by a vote of 99 to 1 (Perrigo, 2025). Ninety-nine senators, in a deregulatory mood, still would not forbid every state in the union from protecting its own citizens from AI harms for a decade. The impulse behind it keeps coming back. Clear the rules out of the way, and worry about the humans later.
This is the cancer cell talking. Not because growth is evil. Growth is good. But growth that refuses to be questioned, growth that treats every guardrail as an insult, growth for its own sake, is the exact thing Abbey was warning us about.
6. But Europe Does Not Get a Halo#
Now let me turn around and kick the other knee, because the European story has a rot of its own.
Draghi is right. A framework that protects rights can also harden into a moat that protects incumbents and smothers the very startups it claims to nurture. Europe’s deeper problem is not that it values dignity. It is that it has built a continent where new companies with new technology are, in Draghi’s telling, not rising at all (Draghi, 2024). Principles do not pay for compute. Arthur Mensch, the CEO of Mistral, says Europe has to “own and operate” its own AI infrastructure or “the money that is being made will not just go back to the hyperscalers in the US” (Geschwindt, 2025). You cannot eat a fundamental right.
And here is the tell. In April 2026, Europe blinked. Under competitiveness pressure, EU negotiators agreed to postpone the AI Act’s core obligations for high-risk systems, pushing them out to December 2027 and August 2028. The lawmaker who steered the deal explained it in language that could have come from the White House. “We need to promote innovation, support startups and scaleups and make it easier to build AI in Europe” (European Parliament, 2026).
So Europe is not pure. Europe feels the gravity of the growth scoreboard too, and it is bending toward it. I will give them this much. The same April 2026 deal that delayed the rules also added a fresh ban on AI tools that generate non-consensual sexual images and child abuse material (European Parliament, 2026). That is a continent still trying, however unevenly, to keep a human being in the center of the frame.
Bending your principles under pressure is not the same as never having had any. It is just the universal human condition. The Visigoths always sack Rome. The question is whether you still remember what you were guarding.
In Summary#
Pull on the threads and the simple story comes apart.
Is Europe behind on the American scoreboard? Yes. Genuinely, measurably behind, and its own former central banker will tell you so to your face. But Europe was never trying to win that game. It chose a different finish line, privacy and dignity and the flourishing of the person, and it wrote that line into its highest law. Those things are real. They are also maddeningly hard to put on a graph and dangerously easy to mortgage the second the growth numbers turn ugly.
America chose dominance, and it is dominating. That is a real achievement, and I am not going to sneer at it. But a country can win every quarter and still lose the plot, if it forgets that all of the building was supposed to be for someone.
So keep both ideas in your head at once. Innovation is good, and Europe needs far more of it. And innovation is a means, not an end, and America keeps talking as if it forgot. The trap is the word itself. “Innovation” has become a thing we say so that we never have to answer the older question sitting underneath it.
All this power, in service of what.
Growth for the sake of growth is the ideology of the cancer cell. Flourishing for the sake of the human being is the older idea, and the harder one, and the only finish line worth crossing. Whoever remembers that is running the race that actually counts.
References#
Abbey, E. (1977). The journey home: Some words in defense of the American West. Dutton.
AI Action Summit. (2025, February 11). Statement on inclusive and sustainable artificial intelligence for people and the planet. Presidence de la Republique. https://www.elysee.fr/en/emmanuel-macron/2025/02/11/statement-on-inclusive-and-sustainable-artificial-intelligence-for-people-and-the-planet
Browne, R. (2025, September 9). AI firm Mistral valued at $14 billion as chip giant ASML takes major stake. CNBC. https://www.cnbc.com/2025/09/09/ai-firm-mistral-valued-at-14-billion-as-asml-takes-major-stake.html
Charter of Fundamental Rights of the European Union. (2000). Official Journal of the European Communities, 2000/C 364/01. https://www.europarl.europa.eu/charter/pdf/text_en.pdf
Draghi, M. (2024, September 17). Address at the presentation of the report on the future of European competitiveness. European Commission. https://commission.europa.eu/topics/eu-competitiveness/draghi-report_en
European Commission. (2021, April 21). Europe fit for the digital age: Commission proposes new rules and actions for excellence and trust in artificial intelligence (IP/21/1682). https://ec.europa.eu/commission/presscorner/detail/en/ip_21_1682
European Data Protection Board. (2023, May 22). 1.2 billion euro fine for Facebook as a result of EDPB binding decision. https://www.edpb.europa.eu/news/news/2023/12-billion-euro-fine-facebook-result-edpb-binding-decision_en
European Parliament. (2026, April 27). AI Act: deal on simplification measures, ban on “nudifier” apps. https://www.europarl.europa.eu/news/en/press-room/20260427IPR42011/ai-act-deal-on-simplification-measures-ban-on-nudifier-apps
European Union. (2024). Regulation (EU) 2024/1689 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act), Article 1. https://artificialintelligenceact.eu/article/1/
Geschwindt, S. (2025, February 13). Mistral CEO: Europe must ‘own and operate’ its AI infrastructure. The Next Web. https://thenextweb.com/news/mistral-ceo-wants-europe-to-own-ai-infrastructure
Google DeepMind. (n.d.). In Wikipedia. Retrieved June 14, 2026, from https://en.wikipedia.org/wiki/Google_DeepMind
James, L. (2026, June 7). ASML becomes Europe’s most valuable company ever as analysts bet on higher EUV output. Tom’s Hardware. https://www.tomshardware.com/tech-industry/asml-beocmes-europes-most-valuable-company-ever-as-analysts-bet-on-higher-euv-output
Kraut, R. (2022, July 2). Aristotle’s ethics. In Stanford encyclopedia of philosophy. https://plato.stanford.edu/entries/aristotle-ethics/
OECD. (2024, May 3). Recommendation of the Council on artificial intelligence (OECD/LEGAL/0449). https://oecd.ai/en/ai-principles
Perrigo, B. (2025, July 1). Senators reject 10-year ban on state-level AI regulation in blow to Big Tech. Time. https://time.com/7299044/senators-reject-10-year-ban-on-state-level-ai-regulation-in-blow-to-big-tech/
Stanford HAI. (2025, April 7). The 2025 AI Index report. Stanford Institute for Human-Centered AI. https://hai.stanford.edu/ai-index/2025-ai-index-report
The Motley Fool. (2026, May 4). The “Magnificent Seven” have a combined market cap of just under $23 trillion. https://www.fool.com/investing/2026/05/04/the-magnificent-seven-have-a-combined-market-cap-o/
Trump, D. J. (2025, January 23). Removing barriers to American leadership in artificial intelligence (Executive Order No. 14179). Federal Register. https://www.federalregister.gov/documents/2025/01/31/2025-02172/removing-barriers-to-american-leadership-in-artificial-intelligence
Tyson, A., & Kikuchi, E. (2023, August 28). Growing public concern about the role of artificial intelligence in daily life. Pew Research Center. https://www.pewresearch.org/short-reads/2023/08/28/growing-public-concern-about-the-role-of-artificial-intelligence-in-daily-life/
Vance, J. D. (2025, February 11). Remarks at the artificial intelligence action summit in Paris, France [Speech transcript]. The American Presidency Project. https://www.presidency.ucsb.edu/documents/remarks-the-vice-president-the-artificial-intelligence-action-summit-paris-france
White House. (2025, July 23). Winning the race: America’s AI action plan. https://www.whitehouse.gov/wp-content/uploads/2025/07/Americas-AI-Action-Plan.pdf