The U.S. government issued a directive on Friday ordering Anthropic to immediately suspend worldwide access to Claude Fable 5 and Claude Mythos 5, citing national security concerns over an alleged jailbreak of Fable 5. Anthropiccomplied but publicly disputed the government’s characterisation, arguing that the identified vulnerability was a narrow, non-universal issue representing capability levels already available in other public models including OpenAI’s GPT-5.5. Some reports indicated the security concerns were first raised with the government by Amazon chief executive Andy Jassy.
The action came just three days after Fable 5’s public launch and immediately disrupted access for hundreds of millions of users globally. Anthropic warned that applying such a standard across the industry would effectively halt all new frontier model deployments.
The shutdown triggered immediate debate in India, one of Anthropic’s largest markets outside the United States. Founders, investors, and policy experts argued the episode exposed the risks of dependence on foreign AI infrastructure subject to unilateral geopolitical decisions. Zoho founder Sridhar Vembu urged Indian organisations to embrace open-source and smaller models, while former Infosys executive Mohandas Pai called for a national AI mission backed by substantial government investment. Technology policy expert Prasanto Roy described the episode as demonstrating that American AI models are inseparable from American geopolitics, a lesson likely to accelerate sovereign AI efforts worldwide.
