I was at a brief press scrum where Arthur Mensch answered a question about the Pope and didn’t flinch. You could see the choices laid out for Europe — restraint or power — and they felt immediate. The room hummed with a sense that a decision now would shape who writes the rules tomorrow.
At a Paris press line, Mistral’s boss pushed back and made a geopolitical claim
Arthur Mensch told reporters that Europe cannot afford to be disarmed while rivals race ahead. He framed the debate as a matter of survival: if adversaries build and control the most advanced systems, Europe risks being excluded from both the benefits and the safeguards of future AI.
Mistral is not speaking from theory. The company just announced a 10‑megawatt data center near Paris and signed deals with Airbus and BMW. France’s government has already moved away from Microsoft Teams and Zoom toward the domestic Visio platform, and the armed forces have a framework agreement to use Mistral models — small changes that signal a larger strategy.
The industry has become a chessboard of influence, and Mensch made clear he wants Europe to hold more pieces.
Why did the Pope call for AI to be “disarmed”?
Pope Leo XIV’s new encyclical, Magnifica Humanitas, warned that AI can dehumanize people and aggravate inequality, pointing to deepfakes, synthetic companions and battlefield uses. “In the era of artificial intelligence, when human dignity is threatened by new forms of dehumanization, ours is the pressing duty to remain profoundly human,” he wrote.
You can feel the moral urgency: the Pope worries that unchecked systems will erode agency and dignity, and he urged a restraint that prioritizes humans over raw capability.
At a conference table, Mistral’s scientists argued AGI access must be European
Guillaume Lample publicly said AGI or superintelligence is likely coming and Europe needs access to it to avoid being cut off from breakthroughs in medicine and science.
That’s the core of Mistral’s claim: missing the frontier models means missing cures, discoveries, and then losing leverage in the geopolitics of information. Lample urged urgency; Reuters and The Wall Street Journal have recorded the company’s pitch that advanced models could accelerate research in ways current tools cannot.
Brussels has become a pressure cooker of policy: national leaders demand technological autonomy while firms chase global scale, and the tension is boiling over into trade decisions and defense partnerships.
Is Europe falling behind the U.S. in AI?
Short answer: in some dimensions yes, in others not yet. U.S. firms — OpenAI, Google, Anthropic, Microsoft — dominate funding, cloud scale, and integration with consumer products. That concentration gives them reach and influence.
But Europe is responding with funding, regulations, and homegrown alternatives. Mistral’s new data center and agreements with Airbus, BMW and French defense show concrete moves. The European Commission’s tech sovereignty package, expected June 3, is meant to push that further. Meanwhile, Chinese labs release open weights and U.S. frontier labs tend to keep their best systems closed — a split that changes how innovation spreads.
At a podium, the argument shifted from ethics to access
Mensch framed the Pope’s call as noble but potentially impractical if it leaves Europe defenseless or dependent. He argued you can’t pause capability across the board without ceding control to others.
I’m not telling you to choose sides. What I will say is this: the battle over AI is simultaneously moral, economic and strategic. The Pope raises ethical red flags; Mistral warns of strategic abandonment. Both are real, and both demand a public conversation about who decides how these systems are built and used.
What is Mistral doing in response?
Mistral is building infrastructure and signing commercial and defense deals to ensure its models run inside Europe. The company’s moves — the Paris data center, partnerships with Airbus and BMW, and the French armed forces framework — are attempts to create an accessible European alternative to U.S. cloud and AI stacks.
They’re positioning Mistral as a counterweight to OpenAI and Google, while also arguing that restricted access to frontier systems would be worse for public welfare than having domestic control.
I’ve watched these conversations for years, and I can tell you the tension will not vanish: who gets to build, who gets to decide, and who benefits are now the axis of power — which side will you bet on?