France Pledges €655 Million for AI and Drops Palantir for Domestic Rival ChapsVision
Prime Minister Lecornu announced a €655 million AI investment package on June 16 as France hosts the G7, while the country's intelligence agency terminates its Palantir contract in favor of French company ChapsVision. A Mistral-powered chatbot will be deployed to over one million public servants.
France is moving fast on sovereign AI. On June 16, Prime Minister Sébastien Lecornu announced a €655 million ($760 million) additional investment in AI through 2030 — timed to France’s presidency of the G7 summit, currently running June 15–17 on French soil.
The announcement’s sharpest edge is the Palantir break. France’s domestic intelligence service (DGSI) is terminating its contract with the US data-analytics firm and replacing it with ChapsVision, a French company. Lecornu was unambiguous: “We cannot depend on foreign powers for our digital tools.”
France is not alone in rethinking Palantir. Germany’s military and London’s Metropolitan Police are both renegotiating their agreements with the company. The cumulative pressure from multiple European governments suggests Palantir’s expansion strategy on the continent faces a structural headwind that won’t resolve quickly.
The €655 million covers AI infrastructure, computing capacity, research, enterprise adoption, and industrial deployment. That number is separate from a larger headline figure announced simultaneously at the G7: the UAE sovereign fund MGX and French public investment bank Bpifrance committed €7.5 billion to build a new AI campus on French soil. That deal — private capital, not state budget — is what Macron is showcasing to the world this week as proof that France can attract investment without ceding data sovereignty.
On the public-services side, a Mistral-powered AI assistant will be rolled out to over one million public servants across state agencies. The health insurance portal Ameli — used by tens of millions of French citizens for medical records and reimbursements — gets a public health chatbot. Both deployments use domestic models, not American APIs.
Mistral is the clear winner here. The Paris-based lab, which closed a reported €722 million debt round earlier this year to build Nvidia-powered European data centers, now has the French state as a reference customer at meaningful scale. That matters enormously when pitching other sovereign clients — governments from Brazil to Southeast Asia are watching which AI providers European states actually trust.
The move also signals something broader. After years of European tech companies being primarily suppliers to US hyperscalers, the AI wave has given some European players — Mistral above all — genuine leverage with their home governments. French digital sovereignty was a talking point for years. In 2026, it’s a procurement decision.
G7 finance ministers and digital ministers are expected to publish a joint statement on AI governance before the summit closes on June 17. Whether that includes binding standards or just shared principles will determine whether this week was strategy or ceremony.