By Dobos Gábor, leader of the Prométheus Society. This piece appears in Ekonomi, the opinion column of G7; the views expressed are the author’s own.
Some 250 years ago the small English firm Boulton & Watt played a pivotal role in making Europe the world’s technological center. The author argues that today the French company Mistral AI represents the opportunity that could prevent the continent from being relegated to the periphery in the age of artificial intelligence (AI).
Why it matters now
AI will fundamentally reshape the economy: according to the author, we face one of the fastest, deepest technological shifts in history. Automation driven by AI will extend across sectors and will increasingly affect office and service work. Countries and firms that respond with restrictive rules or adopt AI half-heartedly will fall behind in global competition, and workers without AI skills will be disadvantaged in the labor market.
If a handful of the most competitive AI firms dominate a market that represents several percentage points of global GDP and none of those firms is European, then the income, tax and consumption losses resulting from disappearing jobs will accrue outside the EU; data-related benefits will also flow abroad. This is not just a missed growth opportunity but a potentially long-term, percentage-point decline in economic performance.
Productivity estimates
The article cites forecasts on AI-driven productivity gains: Goldman Sachs estimates annual gains of 1.5–2.9 percent, while McKinsey’s estimate can be as high as 3.8 percent per year.
Three myths to abandon
The author identifies three common errors in Europe’s thinking:
- Relying on traditional strengths instead of AI. Europe’s large firms still lead in mature technologies where breakthrough potential is limited. Yet the main innovation engines now are areas like autonomy and batteries in autos and AI-driven drug development—fields where Europe is behind.
- Specializing narrowly is insufficient. Without its own general-purpose AI models, Europe would become dependent on American and Chinese models, paying with profits, data and sovereignty. That dependency would spread to industry-specific applications across auto, finance, telecoms, healthcare, logistics and retail.
- Believing a new entrant can easily break into the global top tier is naive. Performance gaps among leading AI models tend to persist; latecomers such as xAI or leading Chinese firms find it hard to catch up with the OpenAI–Google–Anthropic cluster.
Europe’s current position and Mistral’s role
The author contends that Europe currently has only one frontrunner AI firm: Mistral AI. While its most advanced models are somewhat short of the absolute cutting edge, they rank among the best outside the US and China. No other European company compares.
This concentration is both a disadvantage (less internal competition) and an opportunity: with supportive regulation and investment, Mistral could become dominant within Europe. France’s high share of nuclear energy further strengthens its position to meet the energy demands of data centers, offering a competitive advantage even relative to the United States.
Two prerequisites for success
To make Mistral—and thereby Europe—successful, the article singles out two conditions:
-
Build much larger data centers.
- Last May Bpifrance (the French state investment bank) and MGX (an Emirati state investment fund) set up a joint company with NVIDIA and Mistral. The project envisions an AI hub funded by about €8.5 billion, to be established near Paris by 2028.
- The current plan aims for a 1.4 gigawatt final capacity by 2030, but the author argues that this will only be medium-sized and proposes a more ambitious target: a 5 gigawatt data center by 2030.
- Financing should come from within Europe: France should create mechanisms—such as a pooled financial vehicle—to allow European countries to buy stakes in the project and become invested in Mistral’s success.
- The author favors an intergovernmental framework over slow, bureaucratic EU institutions. He notes the EU’s AI Continent Action Plan exists but criticizes its ambition: the EU’s so-called Gigafactories equate to 0.1–0.2 gigawatts, an order of magnitude smaller than what he deems necessary.
-
Regulation must enable, not impede, tech firms.
- Mistral must operate under the same rules as its American competitors—particularly regarding the creation and management of large integrated datasets—otherwise the company will start the race with its legs tied.
- Simplification is necessary, and authorities must abandon the practice of tying companies in administrative knots under the pretext of precaution. While the US labels its AI plan “Winning the Race,” the EU’s top ambition appears to be making the AI market the most regulated in the world—a stance that, the author warns, risks economic stagnation and a drop in living standards.
- Overzealous enforcement and inconsistent rules push tens of tech firms and founders out of Europe each year, mostly to the United States and partly to the United Arab Emirates. Their creativity then benefits economies outside Europe.
The EU’s regulation should protect the single market and promote Mistral’s competitive edge: access to the EU market is the company’s greatest potential advantage, alongside French nuclear power. Regulation must both enable European firms and reassure European governments to choose a European AI provider.
Risks and the lack of a viable alternative
The author acknowledges the possibility that some assumptions may prove wrong: AI’s impact might be less transformative than predicted, or Mistral might fail to keep pace. Even so, the proposed program would still yield benefits by creating one of the world’s largest data centers and strengthening European innovation capacity.
Conversely, if the scenario plays out as forecast and Europe fails to act, the continent risks further sliding behind technologically and economically. Historically Europe fell behind the United States, then China; the Gulf region could be next. Long-term technological decline translates into economic decline, making it unsustainable to maintain a high standard of living without producing leading high-value goods and services domestically.
Conclusion
The author frames the moment as Europe’s last chance: the world is changing inexorably, and Europe must decide whether it will shape that change or be its victim. The prescription is clear—support Mistral, build large-scale, Europe-financed data centers, and adopt enabling regulation. While this implicitly accepts a stronger French technological position, the alternative—total exclusion from the AI revolution—would be far worse.
