The Trump administration ordered AI firm Anthropic to block foreign users of its newest models, Fable 5 and Mythos 5, on Friday. The directive gave the company 90 minutes to disable access, citing an undefined national‑security risk. Anthropic complied, shutting down the services for all customers, and delayed joint experiments with university labs globally.

The dispute follows months of tension after the Pentagon labeled Anthropic a supply‑chain risk and President Trump publicly vilified the firm. Earlier this month the company filed confidentially for an IPO, valuing it at nearly $1 trillion. Critics say the administration’s “jailbreak” claim—alleging a method to bypass model safeguards—is significantly overstated, noting similar capabilities exist in OpenAI’s latest release.

Tech giants including Amazon reportedly warned the White House about the models’ ability to discover and weaponize vulnerabilities, prompting the export‑control action at 5:21 p.m. Friday. Commerce Secretary Howard Lutnick scheduled further talks with Anthropic, but insiders fear the abrupt shutdown could stall the company’s IPO and signal harsher regulatory scrutiny for other AI developers across the sector in the near term.