Rubio tells diplomats to counter AI ‘kill switch’ fears

TL;DR:

• A State Department cable dated 16 July instructs US diplomats to push back on talk of a “kill switch” in American technology and to sell allies on US AI products.
• It follows Washington’s short-lived June block on foreign access to Anthropic’s Mythos and Fable models, which prompted the firm to suspend them globally.
• European lawmakers cited the episode as proof they cannot build critical systems on access that a foreign government can switch off.

The US is running a diplomatic clean-up after a policy misstep rattled its allies. A cable reviewed by Reuters shows Secretary of State Marco Rubio directing diplomats worldwide to counter the argument that American AI comes with a hidden off-switch — an argument the US itself handed critics last month.

How the alarm started

In June, the Trump administration blocked foreign nationals from Anthropic’s most advanced models, Mythos and Fable, on national security grounds, and the company abruptly suspended global access. The ban was lifted within the month, but the message landed. Paired with an executive order asking AI firms to submit models for 30 days of security testing before release, it signalled that Washington was willing to police — and potentially withdraw — products even from allies at short notice.

European politicians seized on it. One MEP said the US “holds a real ‘kill-switch’ over essential technologies”; another warned the continent “cannot keep building its tech stack on access that can be switched off overnight by a foreign government.” The cable’s rebuttal — that a testing window is “not a ‘Kill Switch’” and there is “no government magic button” — is an attempt to unwind that perception.

Why it matters for the UK

This is the sharp end of the sovereignty debate Britain is already living. UK reliance on US frontier models has been flagged by MPs as a national security risk, and Europe’s scramble for alternatives is visible in deals like Microsoft’s funding of Mistral’s compute build-out. The cable also pushes diplomats to portray home-grown “sovereign AI” efforts as wasteful — precisely the capability many UK and EU policymakers now consider essential.

Looking forward

Rhetoric is unlikely to rebuild trust the ban spent. As one European lawmaker put it, Washington has shown “no indication” it is a reliable partner. The lasting effect may be to accelerate the very digital-sovereignty projects the cable was written to discourage.