Anthropic, the US startup behind the advanced Claude Mythos AI model, is set to brief the Financial Stability Board (FSB) on the implications of its technology, which has raised alarms over potential cyber defence threats. The FSB, chaired by Bank of England Governor Andrew Bailey, will receive insights from Anthropic regarding the model's capabilities.
Mythos AI Model: Capabilities and Concerns
Anthropic has chosen not to release Claude Mythos publicly after discovering that the AI model possesses advanced abilities to identify previously unknown flaws in IT systems, which could be exploited by hackers. Instead, the company has granted access to a select group of tech companies and banks, including Apple and JP Morgan, to help them detect weaknesses that the AI might uncover.
The UK's AI Security Institute (AISI) recently updated its assessment of Mythos after evaluating the version provided to banks and tech firms. The AISI noted a "notable capability jump" compared to the preview version tested the previous month. Specifically, the latest iteration of Mythos successfully completed a previously unsolved cybersecurity test called "cooling tower" in three out of ten attempts—a first for any model assessed by the AISI.
AISI Observations and Future Testing
According to the AISI, "Frontier AI's autonomous cyber and software capability is advancing quickly: the length of cyber tasks that frontier models can complete autonomously has doubled on the order of months, not years." The institute is developing new, more challenging hacking tests to keep pace with AI models' progress.
Global Financial Stability Implications
The FSB, which monitors and recommends measures for the global financial system, includes officials from leading economies such as the US, UK, Australia, and China. Its steering committee comprises senior central bank and finance ministry officials. An FSB spokesperson confirmed the engagement, stating the body "welcomes engagement with Anthropic and other firms on emerging and frontier risks to global stability."
The International Monetary Fund (IMF) recently highlighted that financial stability risks are rising due to "fast-moving" developments in AI, calling for a coordinated response. "Cyber risk does not respect borders. As AI capabilities spread across countries, inconsistent oversight could weaken a globally interconnected system," the IMF said in a blog post.
Industry Leaders' Perspectives
Goldman Sachs CEO David Solomon remarked that he is "hyper-aware" of Mythos's capabilities, while JP Morgan CEO Jamie Dimon noted that AI has made cyber defence "harder" but could ultimately help companies defend against hackers. Some experts temper fears, arguing that Mythos represents an evolution rather than a revolution in cyber threats, as most breaches still stem from well-established risks like weak authentication and unpatched vulnerabilities.
Regulatory Response and Guidance
Financial Conduct Authority (FCA) CEO Nikhil Rathi, speaking at the City Week conference in London, confirmed that AI developments were a significant topic at recent IMF meetings. He noted that Bailey is "engaged" on the issue and that cooperation with US authorities is ongoing. Rathi also pointed to guidance released by UK regulators and the Treasury, directing firms to "double down" on "core cyber hygiene," including reviewing legacy systems, implementing good detection mechanisms, ensuring robust governance, and planning for recovery and insurance.
Rathi added that "Anthropic has acted fairly responsibly in engaging with authorities" internationally on the risks of its AI. Anthropic has been approached for comment.
