Anthropic shuts down advanced AI models after U.S. security order
Story Timeline
8 days · 2 summary articles
Anthropic shuts down advanced AI models after U.S. security order
Anthropic urges global pause on cutting-edge AI development amid control fears
Follow-uprevised 3×
The Trump administration ordered Anthropic to suspend foreign access to its two most advanced AI models, Fable 5 and Mythos 5, on Friday, citing national security concerns, prompting the company to disable the systems entirely rather than comply with the directive. The US government’s export control order, issued on 13 June 2026, alleged vulnerabilities in both models that could enable cyberattacks or biological weapon development, claims Anthropic disputes as already addressed in prior testing . With no feasible way to restrict access to non-US citizens without global shutdowns, Anthropic chose to deactivate the models for all users within hours of receiving the order.
Mythos 5, described by sources as capable of identifying software flaws and designing pathogens, was released alongside Fable 5, a more restricted variant intended for public use . Both were pulled from service after regulators warned that even Fable 5’s safeguards could be circumvented through “jailbreak” techniques. The abrupt deactivation marks an unprecedented intervention by Washington into a leading US AI firm’s product rollout, occurring just three days after the models’ public debut.
International media outlets reported the shutdown as a direct consequence of escalating tensions between Silicon Valley and the Trump administration over technology export controls designed to maintain America’s lead over China . Anthropic’s chief executive acknowledged the decision in a statement late Friday, calling the government’s demands technically unworkable and warning that the move could set a precedent for broader restrictions on AI innovation. Legal experts noted that the case may hinge on whether the alleged risks qualify as imminent threats under existing export regulations, a threshold critics argue has not been met.
Analysts suggest the episode underscores the fragility of the US AI sector’s global dominance, which relies heavily on semiconductor export bans that prevent Chinese firms from accessing equivalent hardware . While Washington frames the action as a defensive measure, industry observers warn that unilateral curbs risk accelerating rival AI development elsewhere and could push international customers toward non-US alternatives. The models remain offline as of Saturday morning, with no timeline for their potential reinstatement.
- 3
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- 1
Articles
Anthropic suspende el acceso público a su IA más avanzada por orden de EE.UU. La tecnológica explicó que la Administración de Trump le obligaba a impedir el acceso este nuevo servicio a personas extranjeras
la vanguardia · about 2 hours ago
Anthropic disables Fable 5, Mythos 5 after US security directive Anthropic announced on Friday that it is disabling its Fable 5 and Mythos 5 artificial intelligence models for all users after receiving a US government export control directive citing national security concerns, arguing that the alleged safety vulnerabilities were minor and already addressed in testing.
yenisafak · about 2 hours ago
Anthropic a închis accesul la ultimele sale modele de inteligență arificială, Fable 5 și Mythos, la cererea Guvernului SUA SUA au ordonat companiei Anthropic să suspende accesul cetățenilor străini la ultimele sale modele de inteligență arificială Fable 5 și Mythos 5. Guvernul american invocă „motive de securitate națională, relatează Reuters.
digi24 · about 2 hours ago
Künstliche Intelligenz: Anthropic sperrt Top-KI-Modell nach Regierungs-Anordnung
die zeit · about 2 hours ago
Anthropic sperrt KI-Modell nach Regierungsanordnung
orf.at · about 2 hours ago
Anthropic moet geavanceerde AI-modellen uitschakelen voor niet-Amerikanen Het Amerikaanse AI-bedrijf Anthropic heeft aangekondigd dat het zijn meest geavanceerde modellen "abrupt zal uitschakelen" voor alle gebruikers Dat is een reactie op een bevel van de Amerikaanse overheid om de toegang tot de modellen op te schorten voor buitenlanders.
nu.nl · about 2 hours ago
SUA au cerut Anthropic să suspende accesul cetățenilor străini la ultimele sale modele de inteligență artificială SUA au ordonat companiei Anthropic să suspende accesul cetățenilor străini la ultimele sale modele de inteligență arificială Fable 5 și Mythos 5. Guvernul american invocă „motive de securitate națională, relatează Reuters.
digi24 · about 2 hours ago
USA stoppar Anthropics nya AI-modeller Anthropic har under fredagen stängt av tillgången till två av sina AI-modeller efter direktiv från USA.Beslutet, som förbjuder alla utländska medborgare från att använda tekniken av nationella säkerhetsskäl, sker mitt under en infekterad rättslig strid mellan företaget och presidentstyret.
svenska dagbladet · about 2 hours ago
À lère de lIA, lintelligence humaine est-elle encore spéciale ? Les prouesses de lIA bousculent notre idée de lintelligence, sans effacer la singularité humaine, affirme Tom Griffiths, professeur à luniversité de Princeton, pour The Guardian. Chaque semaine, Courrier international vous propose un billet qui soulève des interrogations sur notre condition moderne en sappuyant sur des œuvres littéraires, scientifiques et, bien sûr, philosophiques.
courrier international · about 2 hours ago
Anthropic suspende el acceso a su modelo más avanzado de IA por motivos de seguridad nacional de EEUU Según Anthropic, para poder cumplir con la directiva de última hora que pide evitar su acceso fuera de Estados Unidos o por parte de extranjeros dentro del país han optado por cortar todo acceso a Fable 5 y Mythos 5 a todos sus clientes Leer
elmundo · about 2 hours ago
Thomas Dekeyser im Gespräch: „Technologien wie Künstliche Intelligenz werden uns aufgezwungen Muss man technologischen Wandel einfach akzeptieren? Ein Interview mit dem Humangeografen Thomas Dekeyser, der offenen Widerstand gegen Innovationen fordert, die von einer Elite kontrolliert werden.
faz · about 2 hours ago
❗️🇺🇸🤖 - Anthropic suspends all access to its Fable 5 and Mythos 5 artificial intelligence models at the Trump administration order on the ground of national security. Mythos 5 is reported to be capable of finding and exploiting cyber vulnerabilities as well as developing biological weapons. Anthropic released Fable 5, a modified version of the former with more guardrails, to the public. However, Fable 5 too was pulled from public access over concerns of jailbreak. US AI companies, particularly Anthropic, hold a significant lead on China thanks to the chip and chip making equipment export restrictions imposed by the US.
telegram bot · about 2 hours ago
Washington invoque la sécurité nationale et force Anthropic à suspendre son IA la plus puissante Le fleuron américain de lintelligence artificielle Anthropic a été contraint vendredi par son propre gouvernement de couper laccès à ses modèles les plus puissants, trois jours après leur lancement, Washington invoquant un risque pour la sécurité nationale dans une décision sans précédent connu. ...
la libre · about 2 hours ago
Künstliche Intelligenz: Anthropic sperrt Top-KI-Modell nach Regierungs-Anordnung So etwas gab es noch nicht: Der OpenAI-Rivale Anthropic muss den Zugang zu seinem besten KI-Modell plötzlich blockieren, weil die US-Regierung ein Sicherheitsrisiko befürchtet. Vieles ist noch unklar.
tagesspiegel · about 2 hours ago
Anthropic « désactive brutalement » ses deux modèles dIA les plus puissants, faute de pouvoir se conformer aux ordres de ladministration Trump Invoquant des questions de « sécurité nationale », Washington exige que la start-up américaine dintelligence artificielle restreigne laccès de Claude Fable 5 et de Mythos 5 aux seuls ressortissants américains. Une demande impossible à mettre en œuvre, selon Anthropic.
le monde · about 2 hours ago