A Close Ally Just Ditched U.S. AI

Palantir Technologies banner displayed on the New York Stock Exchange building

France’s spy chief just fired a warning shot at American tech dominance by moving off a key U.S. AI platform used to hunt terrorists.

Story Snapshot

France’s domestic spy agency is dropping US AI firm Palantir for French rival ChapsVision to “end strategic dependency.”^[2]
Paris is pouring €655 million into homegrown artificial intelligence to escape U.S. control over critical digital tools.^[6]
The trigger was Washington cutting foreign access to a top American AI model, proving how fast the “off switch” can be flipped from D.C.^[6]
Palantir says its French contract is still “fully in force,” meaning a slow, messy transition instead of a clean break.^[3]

France Dumps US Spy Software Over Fears Of American Leverage

French Prime Minister Sébastien Lecornu announced that France’s internal security agency will terminate its contract with American data firm Palantir and move to French company ChapsVision for mass data analysis.^[2] He said France “cannot rely on tools developed by foreign powers” and must avoid “new strategic dependencies in the digital sphere,” making clear this is about sovereignty, not software features.^[5] France’s leaders want critical intelligence systems under French law and French political control.

This decision came just days after Washington cut off foreign access to Anthropic’s advanced Fable artificial intelligence model, jolting European capitals.^[6] That move showed Europe that access to key American technology can disappear overnight if U.S. national security officials decide so.^[6] Lecornu pointed to this as proof France must not depend on the “goodwill” of partners who can “turn off the access tap” for artificial intelligence tools central to security and government operations.^[4]

A €655 Million AI Sovereignty Push That Should Alarm US Policymakers

Lecornu paired the Palantir switch with a plan to invest €655 million in French artificial intelligence infrastructure, including computing power, research, and startups.^[6] The government also wants a single French-built chatbot to serve all state agencies, plus more homegrown tools in health and other sectors.^[2] This is industrial policy wrapped in security language: use state money to build national tech champions and push American vendors out of sensitive systems whenever a local option can be made to work.

ChapsVision, founded in 2019, will now become the “technological foundation” for many French public bodies that need large-scale data analysis, including the spy agency that once depended on Palantir after the 2015 terror attacks.^[3] Germany’s internal security service has already chosen the same French platform, showing this is not a one-off move but a broader European pattern.^[3] Europe’s largest economies are quietly writing “European legal control and data residency” into their procurement rules for intelligence and critical infrastructure.^[14]

Palantir Fights Back, Exposes The Gap Between Rhetoric And Reality

While headlines say France “ended” its Palantir deal, the company insists the long-term contract “remains fully in force” and that it will keep supporting the French government where needed.^[3] Reporting states France renewed Palantir for three more years at the end of 2025 as a bridge until a sovereign tool could fully take over.^[14] Lecornu’s office has admitted Palantir will stay online during a transition to avoid any capability gap in counterterror and counterintelligence work.^[3]

This gap between political messaging and practical reality matters. Publicly, French leaders present a clean break from a controversial American surveillance company co-founded by Trump ally Peter Thiel.^[3] Quietly, they are staging a phased handover because ChapsVision still has to prove it can match Palantir’s performance at scale.^[14] For conservatives in the United States, that should sound familiar: elite politicians posture for their base while relying on American capability when things get serious.

What It Means For America, Trump’s AI Agenda, And US Security

European attacks on “dependence” on American tech should prompt careful thinking in Washington, not panic. On one hand, allies moving to local tools reduces U.S. leverage, data access, and influence over how Western intelligence fights terrorism and hostile regimes. On the other, it exposes how freely Europe will weaponize fears of American spying to justify old-school protectionism and new subsidies for their own firms.^[6] That is the same playbook used for trade, energy, and climate rules for years.

– Rackspace rallies on new AMD partnership and restructuring news. – France ditches Palantir for domestic AI providers to boost strategic autonomy. – Andersen Group beats earnings estimates but shares stay flat.

— Fintwit.ai (@FintwitAi) June 16, 2026

For Trump’s second-term team, the lesson is clear: double down on true American artificial intelligence strength, but do it in ways that respect national sovereignty and the rule of law, including our own Constitution. If Washington uses export controls or model cutoffs too broadly, it will push even close allies into building parallel systems that shut out U.S. companies and reduce shared intelligence. If the administration builds trusted, secure, on-premise deployments and clear legal safeguards, it can keep Palantir and other U.S. innovators at the center of the free world’s security stack without feeding Europe’s anti-American talking points.