Washington DC, United States
Finance ministers, central bank governors and top global bankers have raised serious concerns over a powerful new artificial intelligence model known as “Mythos,” warning that it could pose risks to global financial security systems.
The model, developed by AI company Anthropic under its Claude series, has triggered emergency discussions after reports that it is capable of identifying vulnerabilities across major operating systems and web browsers.
According to experts, the system shows an unusually strong ability to detect and potentially exploit cybersecurity weaknesses, raising fears that it could be misused to target critical financial infrastructure.
Speaking to the BBC, Canada’s Finance Minister, François-Philippe Champagne, said the issue was widely discussed among his counterparts during this week’s International Monetary Fund (IMF) meeting in Washington DC.
He described the situation as serious enough to demand urgent global attention from financial leaders.
Champagne compared the challenge to a “known” risk like the Strait of Hormuz, stressing that unlike that physical chokepoint, the risks tied to Mythos are less predictable and harder to fully understand.
He said:
“Certainly it is serious enough to warrant the attention of all the finance ministers…”
He added:
“The difference with the Strait of Hormuz is that we know where it is and we know how large it is. The issue that we’re facing with Anthropic is that it’s an unknown, unknown.”
Champagne further stressed the need for safeguards and coordinated systems to protect the resilience of global financial networks.
In response to the concerns, top bankers are expected to be given early access to the model so they can test their systems and identify weaknesses before public release.
Barclays CEO, CS Venkatakrishnan, also told the BBC that the development should not be taken lightly.
“It’s serious enough that people have to worry,” he said, adding that financial institutions must quickly understand the risks and fix exposed vulnerabilities.
He noted that the financial sector is moving into a more connected digital environment, which brings both new opportunities and new risks.
Meanwhile, Bank of England Governor Andrew Bailey said the situation requires close attention, especially regarding cybercrime threats.
He explained that advancements in AI are making it easier to detect weaknesses in core IT systems, which could then be exploited by cybercriminals.
“There is a development of AI, of modelling, which makes it easier to detect existing vulnerabilities in sort of core IT systems, and then obviously cyber criminals that the bad actors could seek to exploit them,” Bailey said.
The US Treasury also confirmed it has raised the issue with major banks and is encouraging them to test their systems before any public release of the Mythos model by Anthropic.
Financial industry sources further suggest that another major US-based AI company may soon release a similarly powerful model, though reportedly without the same level of safeguards.