The Trump administration on Friday lifted a two-week-old ban on Anthropic’s advanced AI model *Mythos 5*, allowing more than 100 trusted US cybersecurity firms and government agencies to regain access after weeks of negotiations. The Commerce Department had abruptly suspended the model’s distribution on 12 June, citing national security concerns following the discovery of unspecified vulnerabilities. By Friday evening, the White House confirmed the partial restoration in a statement to *Wired* , easing tensions between Washington and one of Silicon Valley’s most prominent AI labs.

The decision follows a parallel move by the administration to restrict OpenAI’s new GPT-5.6 model to a vetted group of US companies, also announced on Friday. OpenAI confirmed the limited preview in San Francisco and London, describing GPT-5.6 as its most powerful cybersecurity-focused release to date . The White House, citing security reviews, has retained final say over which organisations may use the model, a level of state control over AI diffusion unprecedented outside classified programmes.

European regulators, meanwhile, moved on Friday to tighten oversight of US cloud giants. The European Commission took a preliminary step to designate Amazon Web Services and Microsoft Azure as “gatekeepers” under the Digital Markets Act, despite neither meeting the standard thresholds for dominance. EU tech chief Henna Virkkunen told *Yeni Şafak* that Brussels would enforce “fair, open and competitive” cloud markets .

Anthropic’s partial reprieve signals a fragile détente between Washington and the AI industry. The Commerce Department had forced the company to sever access to *Mythos 5* and its companion model *Claude 4* on 12 June, prompting immediate industry concern over ad hoc regulation. By Friday, the administration had negotiated a narrow pathway for trusted partners, according to *Le Monde* .

The dual announcements underscore a widening global split over AI governance. While the US tightens state-led controls, the EU advances structural market interventions. The juxtaposition highlights the accelerating race for digital sovereignty, with Washington asserting security primacy and Brussels prioritising competitive fairness.

Bluesky Threads

Source Intelligence

20 sources13 countries

Geographic Origin17 located

3 further sources not geolocated

Political Spectrum15 mapped

Articles

OpenAI: US-Regierung lässt Zugang zu neuem KI-Modell von OpenAI beschränken OpenAI erlaubt nur einer kleinen Gruppe von US-Unternehmen den Zugriff auf das neuste KI-Modell GPT-5.6. Grund seien Sicherheitsbedenken der US-Regierung.

die zeit · about 2 hours ago

Inteligența artificială schimbă piața muncii din România. Cum se transformă joburile și ce competențe vor conta tot mai mult Inteligența artificială începe să modifice vizibil modul în care se lucrează în România, nu atât prin dispariția bruscă a unor meserii, cât prin automatizarea treptată a sarcinilor repetitive din tot mai multe domenii, de la contabilitate și resurse umane până la servicii de suport și marketing

adevarul · about 2 hours ago

Ook ChatGPT-maker beperkt toegang tot nieuwe AI op verzoek regering VS, krachtig AI-model Mythos 5 keert beperkt terug

de morgen · about 3 hours ago

EU moves to bring Amazon, Microsoft cloud under tougher rules The European Commission has taken a preliminary step to designate Amazon Web Services and Microsoft Azure as gatekeepers under the Digital Markets Act, citing their dominant positions in the EU cloud market despite not meeting standard thresholds. EU tech chief Henna Virkkunen said cloud services must operate in fair, open and competitive markets.

yenisafak · about 3 hours ago

IA : ladministration Trump autorise Anthropic à réactiver son modèle Mythos pour un groupe restreint de partenaires américains Le ministère du commerce avait brutalement contraint Anthropic à couper laccès à ces deux modèles de pointe, le 12 juin, invoquant la sécurité nationale après la détection de failles.

le monde · about 3 hours ago

Wie is de man achter de gewelddadige AI-videos over Rob Jetten? En wat is er tegen zoiets te doen?

de volkskrant · about 3 hours ago

En Californie, la tribu grandissante des employés de la tech au chômage Le domaine de la technologie subit de plein fouet lirruption de lintelligence artificielle, qui se traduit par des licenciements massifs. Confrontés à la dégradation du marché de lemploi, les travailleurs du secteur congédiés ces derniers mois repensent leur avenir entre retraite anticipée, reconversion ou nécessité détoffer leur carnet dadresses, raconte The Los Angeles Times.

courrier international · about 3 hours ago

Die Sichtbarkeit aller Bevölkerungsgruppen im Internet [premium] Der Wirtschaftsinformatiker Khalid Durani erkundet an der Uni Innsbruck, wie eine Gesellschaft mit KI und Algorithmen gerechter werden kann. Es geht um die Sicherstellung einer verantwortungsvollen digitalen Zukunft.

die presse · about 3 hours ago

European Commission considers classifying AWS and Azure as gatekeepers under Digital Markets Act

gdeltproject.org · about 3 hours ago

USA lubas piiratud juurdepääsu Anthropicu tehisaru mudelile Mythos 5 Ettevõte Anthropic teatas reedel, et sai USA valitsuselt loa anda väikesele rühmale USA küberturvafirmadele juurdepääs oma võimsale tehisaru mudelile Mythos 5.

postimees · about 3 hours ago

Anthropics AI-modell tillåts av USA Anthropic har fått tillåtelse av USA att ge en begränsad grupp av cybersäkerhetsföretag tillgång till sin AI-modell Mythos 5.

svenska dagbladet · about 4 hours ago

Generative AI adoption surges among French mid-sized companies Generative AI use is spreading quickly among French mid-sized companies, but most have yet to report clear productivity benefits, according to a survey by ‌state-backed investment bank Bpifrance. Bpifrance said in its annual barometer of Entreprises de Taille Intermédiaire, or ETIs, that 77 per cent of 534 company heads surveyed said their firms use generative […]

cyprus mail · about 4 hours ago

KI: Berichte – US-Regierung erlaubt Anthropic Weitergabe von KI-Modell an ausgewählte Firmen Nach einer zweiwöchigen Sperre lockert die US-Regierung ihre Beschränkungen für Anthropic. Mehr als 100 ausgewählte Unternehmen erhalten wieder Zugang zu einem leistungsstarken KI-Modell.

handelsblatt · about 5 hours ago

Scrypto Horizontaal: 4. Wat Gooise types in het haar doen, wordt als minderwaardig beschouwd (9) 7. Getransformeerd iemand die je onder handen neemt, en dat vindt veel navolging (11)…

nrc · about 5 hours ago

Así influye el efecto Ricitos de Oro en sus opciones de teletrabajo Lograr un teletrabajo eficaz no es cuestión de azar, sino de diseño. Aplicando el efecto Ricitos de Oro -que transforma el empleo- las empresas descubren que el modelo híbrido coordinado reduce la rotación y mejora el bienestar. La oficina recupera su valor social mientras el trabajo en casa garantiza la actividad profunda. El éxito reside en el equilibrio justo, lejos de los extremos rígidos. Leer

expansion · about 5 hours ago

Digital sovereignty at the UN: Inside the global push to replace US cloud giants with open-source tech

gdeltproject.org · about 6 hours ago

Trump administration allows some access to Anthropics Mythos Move eases tension with AI lab but unease over Washingtons ad hoc regulatory approach remains

financial times · about 6 hours ago

Trump Administration Allows Anthropic to Release Mythos to Select US Organizations After weeks of negotiations, the White House permitted Anthropic to grant access to its most advanced AI model to a select group of US companies and government agencies.

wired.com · about 6 hours ago

Cómo WhatsApp agiliza la comunicación con los hoteles WhatsApp es el principal medio de comunicación entre los huéspedes y la recepción de un hotel. Este sistema es inmediato, natural, eficiente y ofrece una experiencia más cercana al cliente. Leer

expansion · about 7 hours ago

Künstliche Intelligenz: US-Regierung lässt Zugang zu neuer KI von OpenAI beschränken

die zeit · about 7 hours ago

US allows Anthropic to release Mythos to trusted partners reut.rs/44wwCYL

bluesky bot · about 7 hours ago

USA tillåter Anthropics AI-modell

svenska dagbladet · about 7 hours ago

OpenAI präsentiert neues KI-Modell

orf.at · about 7 hours ago

Instagram, filtre și efectul halo. De ce frumusețea valorează mai mult ca niciodată De ce îi percepem pe oamenii frumoși drept mai inteligenți, mai competenți sau mai demni de încredere? Specialiștii numesc acest fenomen „efectul halo, iar cercetările arată că poate influența inclusiv decizii pe care le considerăm obiective.

adevarul · about 7 hours ago

OpenAI apresenta novas versões de IA e começa por parceiros A Anthropic foi a primeira startup de IA a limitar o acesso dos seus modelos a um conjunto de parceiros

observador · about 8 hours ago

OpenAI: OpenAI präsentiert neues KI-Modell für ausgewählte Kunden in den USA Auf Wunsch der US-Regierung präsentiert OpenAI sein neues Modell GPT-5.6 zunächst nur einer begrenzten Gruppe von US-Unternehmen. Grund seien Sicherheitsbedenken.

die zeit · about 8 hours ago