TL;DR
At a June 17 G7 working lunch in Évian-les-Bains, European leaders met AI chiefs Dario Amodei, Demis Hassabis and Sam Altman days after a U.S. export-control order forced Anthropic to block access to its top models worldwide. The meeting sharpened Europe’s demand for durable access, trusted-partner rules and more control over AI infrastructure.
European leaders pressed the heads of Anthropic, Google DeepMind and OpenAI at a June 17 G7 working lunch in Évian-les-Bains, France, after a U.S. export-control directive forced Anthropic to cut off access to its most capable models worldwide, exposing Europe’s dependence on AI systems controlled under U.S. government authority.
The meeting, hosted by French President Emmanuel Macron during the June 15–17 G7 summit, brought Dario Amodei, Demis Hassabis and Sam Altman to the same table as heads of state. According to the source material, the official agenda focused on safe, rapid and effective AI deployment, but the immediate issue was access: whether European firms and public bodies can rely on frontier models that may be restricted by a foreign government order.
The trigger was a June 12 U.S. Commerce Department export-control directive ordering Anthropic to block its Fable 5 and Mythos 5 models for any “foreign national.” Because nationality could not be checked reliably at API scale, Anthropic shut access globally, according to the source material. European organizations using the models lost access without lead time or a phaseout period.
The source material says the lunch also included other technology leaders, including Marc Benioff of Salesforce and Meta AI chief Alexandr Wang, as well as allied and European AI companies such as Mistral, Synthesia, Black Forest Labs, Domyn and Sakana AI. U.S. President Donald Trump attended with senior cabinet officials, while European participants included Macron, German Chancellor Friedrich Merz, European Commission President Ursula von der Leyen and U.K. Prime Minister Keir Starmer.
Évian and the fallout: what Europe actually wants
For the first time, Amodei, Hassabis, and Altman sat with heads of state — five days after Washington switched Anthropic’s models off worldwide. Europe’s question: can you rely on models a foreign cabinet can shut down by decree?
The dilemma: what Europe wants from the three CEOs, the three can’t deliver — because they don’t hold the switch, Washington does. Macron’s platform is the right answer, but no fix for a decade-old infrastructure gap. The only answer that doesn’t depend on someone else’s goodwill: your own models, your own compute, open weights you can self-host.
Europe Wants Model Guarantees
The dispute matters because many companies and public institutions now depend on hosted AI models for daily operations. The Anthropic cutoff turned a policy debate into a service-reliability problem: if an API can disappear by order of a foreign cabinet, businesses face legal, financial and operational risks they cannot fully price.
European leaders want durable access to leading models, protection against another sudden shutdown, a trusted-partner system for non-U.S. allies, and a stronger voice over where compute, chips and power infrastructure are located. The source material also says European officials tied the discussion to child and youth safety, including age rules and protection built into AI services.
Start Here! Learn Microsoft Kinect API
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Shutdown Framed The Summit
Amodei, Hassabis and Altman arrived with broadly aligned proposals. Amodei backed a U.S.-led coalition of democratic states, structured access for trusted partners, chip trade excluding China, and joint defenses against AI risks in cyber operations, biotechnology and intelligence. Hassabis supported a Western coalition and described the moment as one of the most consequential in human history, according to the source material.
Altman called for an international forum to set globally accepted testing standards. The source material quotes him as saying that “no single lab should decide,” a line that captured the executives’ shared message: AI governance should involve governments and democratic institutions, not only the companies building the systems.
Europe’s answer goes beyond access promises. The source material cites a €420 billion sovereignty package, gigafactories and CADA as part of a push for local infrastructure. That effort reflects a wider concern that Europe cannot rely only on U.S.-hosted models, U.S.-controlled chips and U.S. policy choices.
The Model Context Protocol Developer's Handbook: Build, Deploy, and Secure MCP Servers for Claude, GPT, and Local LLMs — The Definitive 2026 Reference … Hardware & Compiler Engineering Series)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Guarantees Remain Unproven
It is not yet clear whether the U.S. directive will be modified, whether Anthropic can restore access for European users, or whether a trusted-partner arrangement can prevent future cutoffs. The source material says the ban remains in place and no reversal has been confirmed.
It is also unclear how much the AI companies can promise on their own. The companies control their products and commercial terms, but the source material’s central tension is that Washington controls the export switch. Any guarantee from executives may be limited unless governments create binding rules.
AI-Powered Cybersecurity: AI Tools for Enterprise Security | AI for Network Security | AI Risk Management | AI in Cyber Policies | Cyber Threat Management AI | ML in Fraud Prevention
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
September Talks Loom
The source material says Western democratic leaders are expected to reconvene in September, while Macron is pushing for a platform within one month. The next test will be whether governments can turn the Évian discussion into enforceable access rules, shared safety standards and infrastructure commitments.
For European users, the practical issue is immediate: deciding how much to rely on hosted U.S. frontier models, and how quickly to invest in alternatives that can be self-hosted or run under European control.
Building AI-Powered Products: The Essential Guide to AI and GenAI Product Management
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What happened at the G7 AI lunch in Évian?
European leaders met AI executives including Dario Amodei, Demis Hassabis and Sam Altman on June 17, 2026, to discuss AI deployment, safety and access after a U.S. directive forced Anthropic to shut off its top models worldwide.
Why did Anthropic shut down access to Fable 5 and Mythos 5?
According to the source material, a U.S. Commerce Department export-control directive ordered Anthropic to block access for any “foreign national.” Because nationality checks were not workable at API scale, Anthropic applied a worldwide cutoff.
What does Europe want from the AI companies?
European leaders want reliable model access, safeguards against sudden shutdowns, trusted-partner rights for allies, stronger local AI infrastructure, a role in compute and chip decisions, and child-safety protections.
Can Amodei, Hassabis and Altman solve the access problem themselves?
Only partly. They can support standards and partner programs, but the source material stresses that the export-control power sits with the U.S. government, not the companies.
What happens after Évian?
Further talks are expected, including a September meeting among Western democracies. The main question is whether leaders can create binding rules before another shutdown tests Europe’s reliance on foreign AI infrastructure.
Source: Thorsten Meyer AI
