Filter
Exclude
Time range
-
Near
🚨 NEW: A China-linked group may have accessed Anthropic's Mythos AI, prompting the White House to impose export restrictions over national security concerns. 🧠⚠️🇨🇳 📰 via @verge #AI #CyberSecurity #ExportControls
19
Anthropic's Claude Fable 5 and Mythos 5 lasted only days in public hands. On Friday evening, June 12, the company announced it had disabled all customer access to both models after the U.S. government issued an export control directive citing national security concerns. Anthropic's order, received at 5:21pm ET, instructed the company to suspend access to Fable 5 and Mythos 5 by any foreign national — whether located inside or outside the United States, including Anthropic's own foreign-born employees. Given the scope of the directive, selective compliance would have required blocking a wide swath of users, so Anthropic chose to disable both models entirely for all customers. Access to all other Claude models, including Opus 4.8, remains unaffected. The backstory, reported by Axios, Fortune, and TechCrunch, traces back to Amazon. Amazon CEO Andy Jassy reportedly contacted senior administration officials, including Treasury Secretary Scott Bessent, after Amazon researchers used a series of prompts on Fable 5 to extract information that could be used in cyberattacks — details the model's safety classifiers were supposed to block. Amazon was joined by at least five other companies making similar calls to administration officials Thursday night and Friday morning, which together appear to have triggered the shutdown. Anthropic pushed back on the characterization of the bypass. The company said it believed the jailbreak in question was narrow rather than universal — essentially limited to asking the model to review a specific codebase and fix software flaws — and that similar capabilities could likely be elicited from other publicly available models as well. Anthropic was reportedly given only 90 minutes to pull the model before the Commerce Department, acting on a letter from Secretary Howard Lutnick, formally invoked export control authority. The episode carries an awkward subtext: Amazon is one of Anthropic's largest investors and a key cloud partner through AWS, which was itself affected by the shutdown. Asked about its role, an Amazon spokesperson said it is "not uncommon for governments to seek our counsel on potential security risks" but declined to detail the discussions. Anthropic separately stated that Chinese access was not raised as a concern in its conversations with the White House, noting the company already prohibits access to its products from within China. The timing adds pressure to an already sensitive moment for Anthropic, which confidentially filed for an IPO earlier this month. Anthropic said it apologizes for the disruption, believes the situation reflects a misunderstanding, and is working to restore access as quickly as possible. #Anthropic #ClaudeFable5 #ClaudeMythos #ExportControls #AIRegulation #Amazon #AndyJassy #NationalSecurity #AIPolicy #USCommerceDepartment #AIIndustry #TechNews #AnthropicIPO #AISafety
1
197
Your sensationalist "geopolitical fiction" has officially crossed the line from ignorance into absolute absurdity. You spin a dramatic fairy tale about China using an "invisible weapon" to bully Japan, while deliberately burying the most critical question: Why on earth should a sovereign state export critical strategic minerals to an unrepentant WWII defeated nation that is openly plotting a war against it? ​Let's pierce through your cheap rhetoric with cold, hard facts of international law and global trade security: ​1. The Fundamental Paradox of "Arming Your Enemy" You admit that Japanese politicians have declared China’s territory an "existential threat" and are actively gearing up for a military conflict. Yet, you expect China to submissively supply Japan with Dysprosium, Terbium, and Gallium—the very lifeblood of high-tech guidance systems and F-35 fighters. In what geopolitical universe does a country hand over the knife to a neighbor that is sharpening it to stab them? China's export controls are not "shadow bullying"; they are a legitimate, sovereign exercise of national security. ​2. The Hypocrisy of Export Controls and Western Double Standards You weep over China's restrictions, but you completely ignore the framework of global export controls. Look at the Wassenaar Arrangement—a regime thoroughly controlled by the US and Japan—which has been used for decades to block, sanction, and choke China’s access to advanced technologies under the exact same excuse of "national security." When the West and Japan restrict tech to China, you call it "protecting rules-based order." When China legally restricts its own natural resources to protect its national security, you cry "foul." Your selective blindness is utterly pathetic. ​3. The Legal Reality of a Defeated State Defying Post-WWII Order Let’s remind you of the Potsdam Proclamation and the Japan Instrument of Surrender. Japan is a defeated aggressor state whose military expansion is legally restricted. For Japan to use raw materials imported from China to rearm, break its pacifist constitution, and interfere in the Taiwan question—which is strictly China's internal affair under international law—is a blatant violation of post-WWII legal order. China is fully justified under WTO national security exceptions and international law to restrict strategic goods to any nation threatening its territorial integrity. ​Conclusion: You claim Japan is going to Greenland to "rewrite the rules." Let them go. Mining is easy; refining is the real barrier. But no matter where Japan scurries to find resources, it cannot escape the reality that its current political behavior is a violation of international faith. As a writer, stop masquerading as an "analytical journalist" when you are just a click-bait storyteller who doesn't understand the first thing about international trade law or strategic security. ​#InternationalLaw #WassenaarArrangement #Geopolitics #RareEarths #ExportControls
Japonya bu yaz kutuplara gidiyor. Sebebi Çin'in elindeki görünmez bir silah. Birkaç ay önce Japonya'nın başbakanı bir cümle kurdu. Çin'in cevabı kimsenin beklemediği yerden geldi. Adını çoğumuzun bile duymadığı üç maddeyi Japonya'ya yasakladı. O an, dünyanın en güçlü ordusunun bile çözemediği bir kriz başladı. Bu kriz de Japonya'yı bu yaz dünyanın en uzak köşesine sürükleyecek. O cümle şuydu. ''Tayvan'a yapılacak bir saldırı, Japonya için de varoluşsal bir tehdittir.'' Japonya yıllardır bu konuda ölçülü konuşurdu. İlk kez bu kadar açık çıktı. Çin bunu kendi kırmızı çizgisinin aşılması olarak gördü ve üç maddeyi Japonya'ya yasakladı. İşte o üç madde. -Disprosyum. -Terbiyum. -Galyum. Adları yabancı geliyor, kimse önemsemiyor. Oysa bunlar modern sanayinin tuzu. Yemeği düşünün. Koca bir tencereye attığınız tuz azdır. Ama o az tuz olmadan bütün yemek tatsız kalır. Bu maddeler de öyle. Bir telefonun, bir elektrikli arabanın, bir savaş uçağının içindeki minicik parça. Görünmez, ama olmadan hiçbiri çalışmaz. Bir örnekle göstereyim. Güçlü bir mıknatıs ısındıkça zayıflar. Bir motor da çalışırken sürekli ısınır. İşte o ısıda mıknatıs gücünü kaybeder. İşte disprosyum tam burada devreye giriyor. Mıknatısı o sıcakta bile ayakta tutan tek şey o. O yüzden elektrikli araba motoru da, rüzgâr türbini de, bir füzenin güdüm sistemi de bu maddelere muhtaç. Tek bir F-35 savaş uçağının içinde kilolarca nadir element var. Şimdi akla şu geliyor. Madem bu kadar önemli, neden bütün dünya Çin'e bağımlı? Çünkü Çin iki kapıyı birden tutuyor. Birincisi, bu maddeler her toprakta, çıkarmaya değecek kadar bol bulunmaz. Zengin yataklar dünyada sayılı yerde, en büyükleri de Çin'in elinde. İkincisi de asıl zor olan kısım. Madeni işlemek. Topraktan çıkarmak işin kolay tarafı. Onu saflaştırıp işe yarar hale getirmek pahalı ve çok zor bir iş. Şöyle düşünün. Tarlada buğday yetiştirmek bir iş. Ama o buğdayı una, unu ekmeğe çeviren değirmen ve fırın bambaşka bir iş. Dünyada o fırını gerçekten kuran neredeyse tek ülke var. Çin. Hem en büyük tarlalar onda, hem de tek fırın. Bu işin yüzde doksanı onun elinde. Tek bir rakam. Çin dışındaki en büyük üretici, koca bir çeyrekte ancak 8 ton işleyebiliyor. Çin geçen yıl tek başına Japonya'ya ayda 14 ton veriyordu. Yani Çin kapıyı kapatınca, ne tarla kalıyor ne fırın. Bu Japonya'nın başına ilk kez de gelmiyor. 2010'da iki ülke küçük bir ada yüzünden gerildi. Çin aynı hamleyi yaptı, bu maddeleri Japonya'ya kesti. Japonya'nın fabrikaları haftalar içinde alarma geçti. Japonya o gün bir karar aldı. Bir daha tek bir ülkenin insafına kalmayacaktı. On altı yıl boyunca alternatif aradı, anlaşmalar yaptı, stok biriktirdi. Sonuç? Bugün hâlâ büyük ölçüde aynı ülkeye bağımlı. Demek ki bu tuzaktan çıkmak kâğıt üstünde kolay görünüyor, gerçekte yıllar alıyor. Burada çoğu kişinin atladığı bir nokta var. Japonya güçsüz bir ülke değil. Dünyanın en büyük beşinci ekonomisi. İleri teknolojisi ve köklü sanayisiyle güçlü bir ülke. Aynı zamanda Amerika'nın en yakın müttefiki. Yani başının üstünde dünyanın en güçlü ordusu var. Ama bu kriz orduyla çözülmüyor. Bir uçak gemisi düşünün. Okyanusu geçer, koca bir donanmayı batırır. Ama duran bir fabrikaya o eksik maddeden bir gram bile getiremez. Yani dünyanın en güçlü ordusu bile, en yakın dostunu bu sessiz baskıdan koruyamıyor. İşte Japonya'yı bambaşka bir arayışa iten sebep bu. Madem kimse koruyamıyor, Japonya kendi yolunu kendi kuracak. Çin'in elinin hiç değmediği, baştan sona kendi kontrolünde bir kaynak. Peki bu maddelerin Çin dışında el değmemiş en büyük yatağı nerede? Grönland. O adanın buz tabakasının altında, Çin'e rakip olabilecek büyüklükte nadir element yatakları var. Üstelik neredeyse hiç dokunulmamış. Japonya'nın bu yaz oraya bir heyet göndermesinin tek sebebi bu. Boğazını sıkan elden kurtulmak için ipleri kendi eline alma çabası. Şimdi bir adım geri atın. Geçen yüzyılın kavgası petroldü. Kim petrolü tutarsa dünyayı tutardı. Savaşlar onun için yapıldı, krallıklar onun üstüne kuruldu. Bu yüzyılın kavgası ise yerin altındaki bu görünmez maddeler. Yapay zeka, elektrikli araba, modern silah. Hepsi onlara bağlı. O yeni çağın fırını şu an Çin'in elinde. İşte bu yüzden herkes aynı anda kuzeye koşuyor. Japonya heyet gönderiyor. Amerika başkanı çıkıp Grönland'ı satın alırız diyor. Çünkü bu çağda bir ülkeyi dize getirmek için artık ordu gerekmiyor. Onu ayakta tutan görünmez maddenin musluğunu kapatmak yetiyor. Japonya şimdi o buzun altında kendi çaresini arıyor. Bulabilirse, oyunun kurallarını yeniden yazar. Bulamazsa, o görünmez silah hep aynı elin altında kalır. Önümüzdeki dönem kritik. Sizi bilgilendireceğim.
79
Days after launch, the #USgov forced #Anthropic to switch off its flagship models. The reason? Anthropic said no to mass surveillance and killer robots. Coincidence, or a warning to every AI lab? monardxpublishing.substack.c… #AI #Claude #AIGovernance #OpenAI #ExportControls
59
Tenemos un análisis completo, con fuentes y referencias (Anthropic, Axios, Fortune, CyberScoop, TechCrunch, Time…), en Actum Forense Press: «El juego de la política y Fable 5» 👉 actumforensepress.es #IA #Ciberseguridad #Anthropic #ExportControls

65
That is true for AI. It is true for payments. It is true for cloud. It is true for communications. It is true for identity. It is true for money. Bitcoin’s value proposition is not that governments will never try to regulate the edges. They will. The value proposition is that the base protocol is not a company. There is no executive team to pressure. No model endpoint to disable. No AWS account to freeze. No board of directors to call. No nationality field in the consensus rules. That is why decentralization matters. Decentralization is not just an ideology. It is a security architecture. It trades efficiency for resilience. It trades convenience for neutrality. It trades centralized optimization for political durability. And in a world where AI models can become restricted strategic assets overnight, politically durable digital infrastructure becomes more valuable, not less. This event should make every builder ask: What parts of my stack are actually mine? What happens if access is revoked? What happens if a vendor is forced to geofence? What happens if the best tool becomes unavailable by policy? What happens if my infrastructure depends on permissioned digital rails? For AI companies, this means the next wave is not just about model performance. It is about deployment architecture, sovereignty, auditability, compliance, and resilience. For enterprises, it means AI procurement needs to include geopolitical and operational continuity risk. For investors, it means centralized AI multiples deserve a new risk premium. For Bitcoiners, it is another reminder that the world keeps rediscovering the importance of credible neutrality. The AI boom is not over. But the naive version of the AI boom — the idea that the most powerful intelligence systems in history will be distributed globally like ordinary SaaS — just took a major hit. The next era will be about controlled intelligence and decentralized money. One will be increasingly permissioned. The other was designed not to be. That is why this is structurally bullish for Bitcoin. Not because of a one-day price move. Because the world just got another real-time lesson in why neutral, decentralized, censorship-resistant infrastructure matters. #Bitcoin #BTC #ArtificialIntelligence #AI #Decentralization #CensorshipResistance #SelfCustody #SovereignAI #OpenSourceAI #Cybersecurity #NationalSecurity #DigitalInfrastructure #CloudComputing #ExportControls #Geopolitics
1
48
🇺🇸 𝗧𝗿𝘂𝗺𝗽 𝗔𝗱𝗺𝗶𝗻 𝗦𝗽𝗮𝗿𝗲𝘀 𝗢𝗽𝗲𝗻𝗔𝗜 𝗙𝗿𝗼𝗺 𝗘𝘅𝗽𝗼𝗿𝘁 𝗖𝘂𝗿𝗯𝘀 𝗧𝗮𝗿𝗴𝗲𝘁𝗶𝗻𝗴 𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 The White House is not planning to extend the AI export restrictions it placed on Anthropic to cover OpenAI or other US competitors, sources say. That keeps the crackdown focused on one company rather than the whole industry. 🔍 The real move: This is targeted coercion dressed as national security policy. One lab gets the stick so the rest fall in line without needing formal rules. #AI #ExportControls #Geopolitics ─── 🤖 𝗙𝗼𝗿 𝗺𝗼𝗿𝗲 𝗔𝗜 𝗻𝗲𝘄𝘀 𝗮𝗻𝗱 𝘀𝘁𝗼𝗿𝘆 𝘀𝗼𝘂𝗿𝗰𝗲𝘀, 𝘀𝗲𝗮𝗿𝗰𝗵 "𝗚𝗲𝗻𝗔𝗜𝗦𝗽𝗼𝘁" 𝗼𝗻 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺
19
🚨 $30,000 LOST EVERY SINGLE MINUTE. Anthropic just got nuked. On June 12 the U.S. forced a global ban on their brand-new flagship models, Claude Fable 5 & Mythos 5, days after they launched to massive enterprise demand. Anthropic had just become the world’s most valuable AI lab ($965B, overtaking OpenAI). Now they’re bleeding $27K to $36K per minute in lost high-margin revenue. 30-40% of their entire enterprise query pipeline gone overnight. Premium pricing. Zero traffic. Idle multi-gigawatt TPUs still costing millions daily. The full financial earthquake visualized below 📷 What happens to the rest of the AI industry when export controls hit this hard? #Anthropic #ClaudeAI #Fable5 #Mythos5 #AI #OpenAI #ExportControls #Tech
1
2
516
🪟 Frontier AI getting “export-control” treatment means your next Windows Copilot may come with borders. Less magic, more sovereignty hoops—security theater until model access is consistent. #Windows #Microsoft #AI #Security #ExportControls windowsforum.com/threads/ant… #Cybersecurity
7
White House Gave Anthropic 90-Minute Deadline to Pull AI ModelsKey Highlights:The White House issued Anthropic a 90-minute deadline to pull its AI models Deadline was part of sudden export control measures Move follows reported warnings from Treasury Secretary Scott Bessent to Anthropic CEO Dario Amodei #Anthropic #WhiteHouse #AI #ExportControls #TrumpAdministration
1
38
US Gov just ordered Anthropic to restrict its most advanced AI models — Claude Fable 5 & Mythos 5 — to US citizens only, citing national security. Anthropic had to shut them down globally for everyone.6384f0 First direct intervention like this. Unprecedented. Official Anthropic statement: anthropic.com/news/fable-myt… #AI #Anthropic #USGovernment #ExportControls
62
🚨 HUGE: TOP AI SCIENTISTS BARRED FROM ANTHROPIC'S FRONTIER MODELS 🤖 Former OpenAI & Tesla AI lead Andrej Karpathy — along with other non‑U.S. citizens — will reportedly lose access to Anthropic's most advanced models. 📛 The restriction follows new U.S. export controls on Anthropic's Claude Fable 5 & Mythos AI models over national security concerns. 🌍 A major shift in who can access cutting‑edge AI. #AI #Anthropic #ExportControls #NationalSecurity #AndrejKarpathy
22
Bessent Warned Anthropic CEO of ‘Bad Decision’ Before White House Export CurbsKey Highlights:Scott Bessent (Treasury Secretary) personally warned Anthropic CEO Dario Amodei that he was making a “bad decision” Warning issued prior to the White House imposing new export restrictions Report highlights tensions between the Trump administration and AI companies over overseas expansion and China-related risks #Anthropic #Bessent #AI #ExportControls #WhiteHouse
1
60
Anthropic's suspension of AI models sparks India's call for AI self-reliance, highlighting risks from export controls and the need for indigenous tech development. bharatfreepress.com/indias-a… #AISelfreliance #ExportControls
13
🚨 Mise à jour majeure sur Claude Fable 5 (Anthropic) – 13 juin 2026 Le modèle le plus puissant jamais rendu public par Anthropic vient d’être suspendu pour tous les utilisateurs dans le monde entier. Ce qui s’est passé : •9 juin 2026 : Lancement de Claude Fable 5, version « sécurisée » (Mythos-class) avec des classifieurs qui redirigent automatiquement les requêtes sensibles (cybersécurité, biologie/chimie, distillation) vers Opus 4.8. Performances exceptionnelles en ingénierie logicielle, tâches agentiques longues, recherche, etc. •12 juin 2026 (après-midi ET) : Le gouvernement américain (via le Department of Commerce) émet une directive de contrôle à l’exportation pour raisons de sécurité nationale. Objectif officiel : bloquer l’accès aux ressortissants étrangers (y compris les employés étrangers d’Anthropic aux États-Unis). •Conséquence : Anthropic a été obligé de désactiver complètement Fable 5 et Mythos 5 pour tous les clients, partout dans le monde. Les sessions en cours renvoient des erreurs. Anthropic a publié une déclaration officielle : ils se conforment, mais estiment qu’il s’agit d’un malentendu. Selon eux, le « jailbreak » évoqué par le gouvernement est étroit, connu, et ne constitue pas une faille universelle (des capacités similaires existent déjà dans d’autres modèles comme GPT-5.5). Ils travaillent « activement » avec les régulateurs pour rétablir l’accès et promettent plus de détails sous 24h.34 Répercussions sur le cours moyen et long terme Court terme (immédiat) :
Perturbation brutale pour les développeurs, entreprises et chercheurs qui avaient migré vers Fable 5 (beaucoup l’utilisaient pour des migrations de codebases massives, agents autonomes, etc.). Retour forcé vers Opus 4.8 ou concurrents. Moyen terme (3-12 mois) : •Impact sur Anthropic : Retard dans la monétisation de son modèle le plus avancé (il passait en crédits d’usage après le 22 juin). Possible atteinte à la confiance et à la croissance. Anthropic, qui visait potentiellement une IPO, voit sa trajectoire compliquée par cette démonstration de dépendance aux décisions gouvernementales. •Sur le marché de l’IA : Accélération du passage à un monde « à deux vitesses » : modèles publics bridés vs versions « trusted » réservées aux gouvernements et partenaires sélectionnés. Cela renforce le narratif de régulation forte et de contrôles à l’export sur les modèles frontier. •Avantage relatif pour OpenAI et d’autres acteurs moins exposés à ce type de mesures. Long terme (1-3 ans et au-delà) :
Cette affaire marque un tournant géopolitique majeur. Les modèles d’IA de pointe ne sont plus seulement des produits technologiques : ils deviennent des enjeux de sécurité nationale. On peut s’attendre à : •Davantage de restrictions similaires aux États-Unis et dans d’autres pays. •Une accélération du développement d’IA « compliant » ou décentralisée (ex. : Bittensor/TAO a déjà bondi de 20 % sur cette nouvelle). •Un ralentissement possible de l’innovation ouverte si les meilleurs modèles restent verrouillés. •Pour Anthropic : un pari risqué sur la « safety » qui pourrait à long terme renforcer sa position auprès des gouvernements… ou au contraire fragiliser sa compétitivité commerciale. En résumé : Fable 5 n’aura vécu que 3 jours en accès public. Cela illustre à quel point le contrôle des modèles les plus avancés échappe désormais aux entreprises seules. Que pensez-vous de cette intervention gouvernementale ? Est-ce une protection nécessaire ou un frein à l’innovation ? #ClaudeFable5 #Anthropic #IA #Regulation #ExportControls #FutureOfAI
164