On 12 June 2026, the United States government ordered Anthropic to deactivate its two most advanced AI models, Mythos 5 and Fable 5, citing a risk to national security. The move, confirmed by Washington on that date, immediately removed the most powerful tools from the global market and exposed Europe’s dependence on foreign AI systems. Analysts in London described the decision as “what Anthropic has sown,” while a German commentator warned that the episode should prompt a “tech turnaround” in the European Union .

The suspension affects the latest versions of Anthropic’s flagship systems, which had been marketed globally just days earlier. European officials privately acknowledge that the bloc lacks comparable high-end models, leaving governments and companies scrambling to secure alternative providers. A senior EU diplomat told *Politico* that Brussels is “wary of stoking” a transatlantic row at the upcoming G7 summit, preferring to frame the issue as a shared challenge rather than a bilateral dispute .

The episode follows a parallel decision in France, where the domestic intelligence service DGSI announced on 17 June 2026 that it will replace Palantir’s data-analysis software—used since the 2015 terror attacks—with a platform developed by the French firm ChapsVision. The transition, described as a “sovereignty choice” by *Le Monde*, is both a political signal and a technical challenge, requiring DGSI to migrate years of operational data to a new system .

Commentators across Europe now argue that AI is not merely a commercial product but a strategic asset whose access is controlled by governments. An editorial in *L’Express* titled the episode “AI and the kids of European sovereignty,” while an Italian newspaper noted that both Washington and Bern had, within 24 hours, clarified that advanced AI is a matter of state policy rather than private enterprise .

With the G7 summit days away, European policymakers face a dual imperative: accelerate domestic AI development and negotiate with Washington to restore access to critical models. The Anthropic suspension has become a case study in the fragility of global AI supply chains and the high stakes of technological sovereignty.

Bluesky Threads

Source Intelligence

6 sources3 countries

Geographic Origin5 located

1 further source not geolocated

Political Spectrum5 mapped

Articles

Usa e Svizzera docent: lAi non è servizio di pubblica utilità, ma una risorsa strategica di proprietà nazionale Nel giro di 24 ore la Svizzera e gli Stati Uniti hanno chiarito una verità che in molti preferivano ignorare: lintelligenza artificiale è materia di Stato. Laccesso ai suoi sistemi più avanzati obbedisce alle logiche di chi comanda — la politica — molto più che al potere dei miliardi. Due diversi episodi, letti insieme, chiudono […] Larticolo Usa e Svizzera docent: lAi non è servizio di pubblica utilità, ma una risorsa strategica di proprietà nazionale proviene da Il Fatto Quotidiano.

il fatto quotidiano · 17 days ago

Des romans contre lURSS : lopération secrète de la CIA pendant la guerre froide

lexpress · 17 days ago

Avec linterdiction de certains de ses modèles dIA, Anthropic récolte ce quelle a semé Invoquant un risque pour la sécurité nationale, Washington a forcé, le 12 juin, Anthropic à désactiver ses modèles dIA les plus puissants, Mythos 5 et Fable 5. Une alerte pour lEurope, dépendante de ces outils. Et une victoire pour la Chine, analyse ce journaliste britannique spécialiste de la tech.

courrier international · 17 days ago

Europe wary of stoking Anthropic row at G7

gdeltproject.org · 17 days ago

Affaire Anthropic : lIA et les cabris de la souveraineté européenne edito Eric

lexpress · 17 days ago

Renseignement : le remplacement de Palantir par le français ChapsVision à la DGSI, un choix de souveraineté au long cours et complexe Utilisé depuis les attentats de 2015, le logiciel danalyse de données de la société américaine controversée doit progressivement laisser la place à celui de léditeur français. Une décision espérée de longue date, mais aussi, un défi technologique.

le monde · 17 days ago

Umstrittene Gen-Eingriffe: Topgenetiker lehnt Design von Embryonen ab Ein Schweizer Spitzenforscher manipuliert präzise menschliche Embryonen in einem US-Labor. Für viele ist es nur ein Test der verwendeten Technik, für den Bonner Genetiker Markus Nöthen Teil eines aufblühenden „Wahnsinns.

faz · 17 days ago

Abhängigkeit und Überregulierung: Europa braucht die Tech-Kehrtwende Die Entscheidung der US-Regierung, den Zugang zu Fable 5 und Mythos 5 auszusetzen, führt vielen Europäern eindrucksvoll unsere Abhängigkeit von ausländischen KI-Spitzenmodellen vor Augen.

faz · 17 days ago