US government orders Anthropic to shut down powerful AI models

The U.S. government has ordered AI firm Anthropic to disable its advanced AI models, Claude Fable 5 and Claude Mythos 5, globally, citing national security concerns. Amazon CEO Andy Jassy reportedly raised security risks associated with these models to Trump administration officials before the order was issued. The U.S. government has restricted access to these models for non-U.S. nationals, which has intensified India's debate on developing its own sovereign AI capabilities and highlighted concerns about dependence on foreign technology.

Amazon and the White House reportedly intervened to end Fable's involvement with AI startup Anthropic. Anthropic was given a 90-minute deadline to remove Fable following a demand from the Trump administration. Cybersecurity researchers are expressing frustration with the restrictive guardrails on Anthropic's new AI model, Fable. The model, intended for cybersecurity applications, frequently rejects even innocuous prompts related to the field, hindering its utility for security professionals.

In a separate development, KPMG has withdrawn its report titled "Redefining excellence in the age of agentic AI" after numerous organizations stated its claims about their AI usage were untrue. The research group GPTZero identified inaccuracies stemming from AI hallucinations, with many citations appearing to be fake. This incident underscores broader concerns about the reliability and accuracy of AI-generated content.