The Trump administration's sudden export controls on Anthropic's most advanced AI models have rattled allied nations and forced the company to disable the technology for everyone, including US users.
The US government just pulled the plug on two of the most advanced AI models in existence, and the timing could not have been more diplomatically awkward.
On June 12-13, the Commerce Department issued export controls on Anthropic’s Fable 5 and Mythos 5 models, citing national security concerns. Anthropic responded on June 13 by disabling the models for all users, including those in the United States, to ensure full compliance. Four days later, world leaders sat down at the G-7 summit in Évian-les-Bains, France, to discuss global AI governance, only to find that the biggest AI story of the week was the US government unilaterally pulling the rug out from under one of its own flagship companies.
What happened and why it matters #
The Commerce Department’s directive required Anthropic to cut off access for all foreign nationals. Rather than risk non-compliance by trying to sort domestic users from international ones in real time, Anthropic chose to shut everything down for everybody.
If you were using Fable 5 or Mythos 5 for anything, whether you sit in San Francisco or São Paulo, you lost access on June 13. President Trump, speaking on June 17 as the G-7 summit got underway, said negotiations with Anthropic were “going fine.” He offered no specific timeline for when the models might be restored.
The G-7 fallout #
French President Emmanuel Macron cautioned that unrestricted, unilateral US control over advanced AI technologies could have adverse economic effects on allied nations. He also noted the potential damage to US businesses themselves.
Both Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman were present at the summit. Several G-7 participants reportedly pushed for a coordinated approach that would allow allied nations to access advanced AI capabilities under shared safety protocols, rather than leaving each country’s access subject to unilateral US decisions.
The bigger picture for AI governance #
Anthropic’s decision to disable the models for everyone rather than attempt granular compliance illustrates the enforcement problem. The compliance burden of verifying every user’s nationality in real time was apparently too high to manage without risking a violation.
What this means for investors #
Anthropic, despite being privately held, sits at the center of a massive web of enterprise partnerships and cloud computing agreements. When its flagship models go dark without warning, every company that built on top of them feels the impact.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our