According to informed sources, AI startup Anthropic is set to brief the Financial Stability Board (FSB) on global financial system network vulnerabilities identified by its latest large AI model, "Mythos." Following a request from Bank of England Governor and FSB Chair Andrew Bailey, the maker of the Claude chatbot will discuss the capabilities of its new Mythos Preview model with key treasury and central bank officials from the FSB. Bailey chairs this global risk oversight body, which coordinates financial regulations for G20 economies. Anthropic describes Mythos, unveiled last month but not yet officially released, as a cybersecurity model designed to detect decades-old vulnerabilities in web browsers, infrastructure, and software. Cybersecurity experts have previously warned that such a system could significantly enhance more sophisticated cyberattacks, posing risks to the banking sector, which relies on legacy technology systems. In April, BoE Governor Bailey cautioned that Mythos could introduce substantial security risks to the cyber world. Speaking at an event at Columbia University in New York last month, Bailey stated, "People would rightly think that the situation in the Gulf is the latest challenge we face in this area. Until—I think it was last Friday—you woke up to find that Anthropic may have found a way to completely break through the entire world of cyber risk." "The core question is: to what extent can this new version of a personal product identify vulnerabilities in other systems that could be used for cyberattack purposes, in a sense," Bailey added. Reportedly, the initial release of Mythos by global AI applications leader Anthropic is not an ordinary AI chatbot but a preview of a cutting-edge, general-purpose large model not yet fully publicly released. Its most prominent capabilities focus on advanced code comprehension, vulnerability discovery, and exploit generation. Anthropic has officially disclosed that Claude Mythos Preview can already discover and exploit "zero-day vulnerabilities" in all major operating systems and mainstream browsers during testing and, in some cases, autonomously generate complex exploit chains with minimal human intervention. Senior financial regulators from multiple countries have recently indicated they are actively monitoring the development trajectory of Anthropic's Mythos, the world's most advanced AI model currently in preliminary testing. Some senior technology experts specializing in the financial sector suggest this large AI model may possess the capability to be used for large-scale disruption of banking system stability. Experts note that Mythos's exceptionally powerful capabilities in high-level agent programming could grant it an unprecedented ability to identify vulnerabilities in cybersecurity systems. This has prompted increased scrutiny from some of the world's most influential policymakers and financial system regulators, such as the Federal Reserve and the Bank of England. These regulators have already begun to view Mythos as a potential core new risk source that could destabilize banking systems, payment clearing, and critical financial databases.
Comments