EVIAN-LES-BAINS (Realist English). The leaders of the Group of Seven (G7), who gathered on June 15 in the French resort town of Evian-les-Bains, will include a special section on artificial intelligence risks to the global financial system in the summit’s final communiqué.
According to Bloomberg, citing a draft of the statement, a clause has been inserted into the document calling for “further discussion of new opportunities and potential risks associated with AI,” with a particular emphasis on threats to the financial sector. Member countries will also urge each other to strengthen coordination in this area.
As noted in the document, Paris and Washington are particularly interested in discussing the technological threat. US officials have already stated that they are prepared to use emergency powers to force AI developers to eliminate potential national security threats. Against this backdrop, the draft statement also includes provisions on protecting the financial sector from risks associated with quantum technologies.
‘Claude Mythos’ as a New Challenge
The discussion of AI risks to the financial sector comes amid the emergence of so-called “advanced” AI, specifically the new powerful neural network “Claude Mythos” from the American startup Anthropic. At a recent meeting of G7 finance ministers and central bank governors in Paris that preceded the current summit, this model was identified as a direct threat to financial stability.
According to TBS, experts fear that “Mythos” could be used by malicious actors for cyberattacks on banking and payment infrastructure. This AI is capable of finding systemic vulnerabilities and hacking the protective mechanisms of digital systems, which, according to experts, could lead to consequences comparable to a serious armed conflict. Japan’s Finance Minister Satsuki Katayama, who chaired the April meeting in Paris, called on G7 countries to develop a joint strategy before the summit.
“It is necessary for the G7 to be able to coordinate its actions when repelling such attacks,” she emphasized.
The Anthropic Case: Making AI Work for Security
As part of preparing the declaration, Washington has already taken a first practical step: at the behest of the US administration, Anthropic, at the request of authorities, banned foreign users from accessing its most advanced model, Fable 5. This was a direct response to concerns that the latest algorithm is being “hacked” to bypass built-in security systems.
As stated in the draft, the G7 members will continue to implement early warning mechanisms and prepare the financial system for the challenges of the new technological era.
In Trump’s Shadow: The Political Background of the Summit
The signing of the AI declaration comes amid behind-the-scenes political battles. According to sources, French President Emmanuel Macron and other EU leaders are doing their utmost to avoid friction in negotiations with Donald Trump on trade and climate issues. The most contentious topics will be omitted from the final document, but the AI section was kept because it represents a mutual interest.
Thus, the G7 is effectively acknowledging that the rapid development of neural networks is becoming a factor of global instability requiring urgent intervention at the level of heads of state.
The final version of the communiqué will be released following the two-day summit, which concludes on June 17.
