G7 leaders explore 'trusted partners' route around US ban on Anthropic AI models
At the G7 summit in Evian-les-Bains, leaders discussed a plan to grant select 'trusted partners' access to advanced AI systems from U.S. firms such as Anthropic, after Washington ordered the company to block foreign nationals from its most capable models.
Background
On 12 June, acting on a US Commerce Department directive, Anthropic blocked foreign nationals from accessing its two most advanced AI models, Mythos 5 and Fable 5. Because the restriction could not be enforced selectively on a shared cloud service, the company disabled the systems for everyone worldwide. The order is believed to be the first export-control measure aimed at specific AI models rather than the chips that run them.
The trusted partner proposal
At the opening dinner of the G7 summit in Evian-les-Bains on 16 June, representatives of several member countries raised with US Commerce Secretary Howard Lutnick the idea of a 'trusted partners' scheme. The partners could be countries or individual companies, according to diplomatic sources. The stated aim is to let G7 allies use the models to strengthen cybersecurity defences against rivals, particularly China.
We have an open line of communication with our allies, and we remain committed to addressing national security concerns with Anthropic's model.
European reaction and UK denial
The ban has intensified European fears of over-reliance on US technology and spurred calls for 'digital sovereignty.' A No. 10 spokesperson denied media reports that the UK had sought a carve-out from the controls, calling them "categorically untrue." The spokesperson said the UK is in "regular and constructive contact" with both the US government and Anthropic. British Prime Minister Keir Starmer, in remarks ahead of the summit, stressed child online safety and warned tech bosses to cooperate.
I don't just want Britain to lead the tech revolution, I want us to drive who this change is actually for. So my message to tech bosses is clear: work with us to keep kids safe or I'll do whatever it takes to make that happen.
AI CEOs at the summit
On 17 June, executives from Anthropic, OpenAI and Google are expected to attend a working lunch with G7 leaders to discuss technology issues including regulation, AI infrastructure and networks. French President Emmanuel Macron is leading talks to explore how to deploy the most advanced models through trusted partners.
- US Commerce Department directs Anthropic to block foreign access to Mythos 5 and Fable 5, forcing a global shutdown.
- G7 dinner in Evian-les-Bains: allied representatives propose a 'trusted partners' scheme to Commerce Secretary Lutnick.
- AI CEOs from Anthropic, OpenAI and Google attend a working lunch with G7 leaders to discuss regulation and infrastructure.
