Three days after the Vatican called for AI to be “incubated,” Mistral CEO defended his company’s defense-AI work, arguing that Europe cannot afford unilateral restrictions.
Arthur Mensch, CEO of French AI startup Mistral, defied Pope Leo XIV’s call on Thursday. “Disarm the AI” It argues that European companies cannot back out of defense-AI work when competitors are actively deploying the technology.
Three days later, the comments made by the Vatican were published Magnificent HumanityThe pope’s first encyclical marks one of the Catholic Church’s most direct corporate responses to what is fast becoming its most influential intervention in artificial intelligence.
“We are all for peace” mensch said “But if you look at our competitors and adversaries around the world, they’re using artificial intelligence. As long as we have adversaries that threaten and they threaten, we have to have our own capabilities.”
💜 of EU technology
The latest rumblings from the EU tech scene, a story from our wise founder Boris and some questionable AI art. Free in your inbox every week. Register now!
The Mistral CEO’s framework is a structural defense of the military-AI development that Europe’s tech sector has been working on since the Ukraine war, but his decision to frame it as an outright rebuttal of the incumbent Pope makes Thursday’s remarks noteworthy.
The encyclical itself is the document to which Mensch responded. Magnificent HumanityThe 42,300-word Leo text, published on May 25, calls for the disarmament of artificial intelligence, the establishment of three mandatory requirements for the deployment of any autonomous weapons, tracking of decisions, meaningful human control of lethal actions and international rules to slow the technological arms race, and clearly rejects conventional “warfare.”
The Pope further argued that military force could only be justified by this “self-defense in the strictest sense.” The encyclical is the most direct papal intervention in tech regulation in decades.
Mensch’s position has its own theological resonance. Pope’s framework of “self-defense in the strictest sense” and Menshin’s framework of “enemies threaten, therefore we need our own capabilities” do not contradict each other.
Both accept the legitimacy of self-defense; both deny abusive use. Their disagreement is over what self-defense requires in 2026. Leo’s position is that the threshold for the introduction of lethal AI is higher than any state has ever expressed.
According to Mensch, Europe cannot meet credible adversaries with this threshold, while these enemies operate without it.
The commercial background is important here. Mistral has been building its defense-AI portfolio since at least early 2025. The Helsing partnership was announced at the Paris AI Action Summit In February 2025, he developed joint work on vision-language-action models designed for “new generation defense systems”.
Helsing has already deployed AI systems in Eurofighter fighter jets, battlefield simulations and Ukrainian drone operations. Mistral has separately applied for defense contracts with several European governments.
Mensch’s public pushback against the Pope is not a hypothetical stance, but a defense of an existing line of business that is now under official moral condemnation by the Vatican.
And the pope’s influence on the policy debate on artificial intelligence has been greater than expected six months ago.
Anthropic co-founder Christopher Olah The document appeared in the presentation of the encyclical, which lent its approval to Silicon Valley. The European Commission welcomed this on Monday evening; OpenAI, Google, and Microsoft have given official expressions of respect.
The Vatican is not, in any meaningful sense, a regulatory body for the development of artificial intelligence. What he has created with Magnifica Humanitas is a moral vocabulary that legislators and politicians can use, and Mensch’s rebuttal acknowledges how important that vocabulary is now in existence.
The pure rhetorical contrast masks the quieter reality of European politics. Brussels is moving towards enforceable AI-warfare frameworks, but has yet to codify the binding limits required by Magnifica Humanitas. Member-state governments are also expanding defense-AI procurement budgets.
The conflict is real, and the next year of EU AI Act implementation, member states’ defense spending and Vatican-aligned policy advocacy will show which side wins.
Mensch chose to bet his company’s public position on the defense-supply side of that argument, according to testimony Thursday.
