Anthropic's path to a nearly $1 trillion（約160兆円） IPO is being tested by repeated government orders to shut down its AI models, raising questions about whether investors are properly accounting for regulatory risk.

What happened: Anthropic took its Fable 5 and Mythos 5 models offline after the Commerce Department barred foreign nationals from using them—the company's second blacklisting by the federal government in six months. The Department of War has also moved at least two-thirds of its AI workflows away from Anthropic's models since the two clashed over military use of Claude, and its secretary publicly declared the move permanent.

Why it matters: For a company filing to go public with a nearly $1 trillion（約160兆円） valuation, the ability of the U.S. government to force products offline overnight represents a form of regulatory risk that could fundamentally reshape the company's revenue potential and future stock price. Some observers expect these standoffs to recur periodically, and foreign customers may shift toward homegrown AI options if Anthropic becomes an unreliable supplier due to government intervention.

What to watch: Reports indicate Amazon—an Anthropic investor with $8 billion（約1.3兆円） already committed and up to $25 billion（約4兆円） more—brought security concerns about Fable to the White House, including the model's tendency to reveal software vulnerability information when questioned. How the company resolves the tension between its largest backer and the federal government will signal whether the valuation can hold through a public offering.