- The ban on Anthropic models triggered a debate about control over the future of artificial intelligence.
- G7 leaders are seeking an alternative to dependence on OpenAI and Anthropic after Washington's decision.
- As a result, countries began talking about technological sovereignty and a "trusted partners" mechanism.
G7 leaders, heads of major AI companies and representatives of the tech industry discussed the risks of overreliance on American artificial intelligence models during a summit in Evian-les-Bains, France. FT reports this.
During the G7 summit, French President Emmanuel Macron and Indian Prime Minister Narendra Modi warned that Washington's ability to cut off access to advanced AI systems at any moment creates risks for the economic and technological security of other states. The situation drew attention to the issue of so-called "digital sovereignty" and the global market's dependence on American AI developers.
The discussion unfolded several days after the administration of US President Donald Trump blocked the export of Anthropic's new models — Claude Mythos 5 and Claude Fable 5 — citing national security considerations.
Earlier, Anthropic had introduced Claude Fable 5 as the most powerful model in its history and opened limited access to Claude Mythos 5 for cyber defenders and biomedical researchers. At the time, the company's chief scientist Boris Cherny called the release the "biggest leap" in the development of Anthropic's models.
Claude Fable 5: testing the first publicly available Mythos-class model 10.06.2026 Read
However, soon after the launch, US authorities required the company to restrict access to the new systems for foreign users both inside and outside the United States. In response, Anthropic temporarily suspended support for the models for all customers, stating that it disagreed with the decision and considered its justification insufficiently convincing.
Risks of dependence on American AI
According to the publication, during a closed-door meeting of G7 leaders and heads of the largest AI companies — among them Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman — Macron urged attention to the consequences of such an approach.
In his words, if the US can "flip the switch from one day to the next," it would harm not only European users but also the American companies themselves, which risk losing the trust of international clients.
For his part, Modi also expressed concern over the decision regarding Anthropic and stressed that democratic countries must have unimpeded access to advanced AI models to protect critical infrastructure.
The incident highlighted a problem that international companies and governments are already facing: any organization that builds its own services on top of American AI infrastructure could potentially lose access to it without warning and without a detailed explanation of the reasons.
An Anthropic customer accidentally spent $500 million on the Claude chatbot in a single month 29.05.2026 Read
During the meeting, Anthropic CEO Dario Amodei urged the G7 countries to "resist the temptation of fragmentation" on the issue of artificial intelligence development.
According to sources, he was supported by Altman. The heads of Anthropic, OpenAI and DeepMind advocated cooperation among democratic states under US leadership in developing advanced AI systems.
Altman also stressed that cyber defense tools should be available to all the countries represented at the meeting.
DeepMind CEO Demis Hassabis also joined the discussion. Together with Amodei and Altman, he emphasized the importance of international cooperation in AI, and also drew attention to the risks of bioterrorism and cybersecurity in the event of fragmentation of democratic alliances.
An AI model from Anthropic "helped" in developing chemical weapons and committing serious crimes 12.02.2026 Read
Against this backdrop, the concept of technological sovereignty is attracting increasing attention — the ability of countries to control critically important digital technologies independently of the decisions of foreign governments.
"The recent restriction of access to Anthropic's models confirms what we at Cohere have known for a long time: the dependence of companies and democratic states on a small group of large tech companies is dangerous for the resilience of systems," said the co-founder and CEO of the Canadian AI company Cohere, Aidan Gomez.
He added that "digital sovereignty is not just a question of competition or the interests of a single company or country." In his view, it concerns those "who control the fundamental technologies that will define our economic security and national sovereignty over the next decades."
G7 discusses creating a "trusted partners" network
Amid the disputes, summit participants also considered the possibility of creating a "Trusted Partners" mechanism that would allow countries outside the US to gain access to advanced models from companies such as Anthropic and OpenAI even amid tightening export restrictions.
According to the preliminary concept, both states and private companies could join the program, provided they use the models to strengthen the technological and cybersecurity of Western allies.
However, it is still unclear how broad such a program would be and whether it could protect startups and businesses from the risk of a sudden loss of access to critically important AI services.
As Macron emphasized, democratic countries should jointly shape the rules for using artificial intelligence.
"What we will build over the coming months is a kind of platform for discussion and cooperation among democratic countries in response to the risks created by artificial intelligence, in order to jointly define common standards," the French president said.
At the same time, the head of the French AI startup Mistral AI, Arthur Mensch, noted that one of the main topics of the meeting was countries' dependence on the global AI technology supply chain.
"When you have an intertwined supply chain, are you sure that your partners cannot cut you off? This question was repeatedly raised, mainly by participants from outside the US," he said.
Notably, the discussion took place amid active calls for tighter control over the development of artificial intelligence from Anthropic's own leadership. Earlier, the company's CEO Dario Amodei said the US should introduce mandatory safety requirements for the most powerful AI models and strengthen government oversight of the industry due to risks to the labor market, security and civil liberties.
Source: Incrypted
Новости в мире криптовалют
Random quote about money
"Деньги — одно из величайших орудий обретения свободы, придуманных человеком."
Interesting posts in other sections of the blog
* to search the proxy database, just enter a country name, e.g. Russia, USA, Thailand