The U.S. government has ordered Anthropic to suspend access to its Fable 5 and Mythos 5 AI models, citing national security concerns over a potential jailbreak. The directive has led to a sell-off in Anthropic's pre-IPO market.
The U.S. government's unprecedented directive to halt access to specific AI models raises significant questions about the balance between national security, technological advancement, and global access to cutting-edge AI. It also impacts the valuation and IPO prospects of leading AI companies and fuels debate about AI self-reliance in other nations.
The U.S. government has ordered AI firm Anthropic to suspend access to its two most advanced AI models, Fable 5 and Mythos 5, citing national security concerns over a potential jailbreak. Anthropic received the directive on Friday evening and has challenged it, arguing the government's evidence of a narrow, previously known vulnerability is insufficient and that similar models, including OpenAI's GPT-5.5, can identify the same issues.
The shutdown order has negatively impacted Anthropic's pre-IPO market, with its perpetual contract on Hyperliquid falling 3.7%. The company stated that applying the government's standard would effectively halt all new model deployments for frontier model providers and plans to release further technical details on the jailbreak within 24 hours.
In India, the U.S. government's action has spurred calls for increased AI self-reliance, domestic research and development, and investment in open-source models, with tech leaders highlighting concerns about structural dependence on foreign AI platforms.