Anthropic AI Mythos: Urgent warning to US banks, BSI expects upheavals
Anthropic's new AI Mythos is causing a stir. In the US, the heads of important banks were summoned; here, the BSI expects far-reaching consequences.
(Image: PhotoGranary02/Shutterstock.com)
US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell convened an unusually urgent meeting with the heads of major US banks on Tuesday to warn of the dangers of Anthropic's new AI model Claude Mythos Preview. This is reported by Bloomberg, citing informed sources. The short-notice meeting is a sign that US regulators consider the possibility of a new type of cyberattack to be among the greatest risks to the financial industry, the financial news platform summarizes. The heads of all banks considered systemically important in the US were summoned. The BSI also expects "upheavals in dealing with security vulnerabilities" from the new AI model and similar successors.
„No more unknown classic security vulnerabilities in the medium term“
Anthropic only presented Mythos on Tuesday and at the same time stated that the model is so dangerous that it will only be made available to companies working on IT security. The AI model has already identified thousands of high-risk zero-day vulnerabilities, including in all major operating systems and every internet browser. At the same time, the AI technology is significantly more capable of developing a working exploit for such vulnerabilities, sometimes even using several in conjunction with each other. As part of "Project Glasswing", the industry is now to work on patching vulnerabilities found by it before other AI models become available with which criminals can also find vulnerabilities much more easily.
Videos by heise
The Federal Office for Information Security (BSI) also assumes that the new technology will have significant implications for the cyber threat landscape, reports dpa. Her authority is in contact with Anthropic, BSI President Claudia Plattner stated upon request. Her agency expects "upheavals in dealing with security vulnerabilities and in the vulnerability landscape as a whole." If thought through consistently, there could be no more unknown classic software vulnerabilities in the medium term, she believes. "This would result in a shift in attack vectors and a paradigm shift with regard to the cyber threat landscape."
Furthermore, the question arises whether and for how long such powerful tools will be available on the open market, dpa further quotes. Given the rapid development of AI technology, it can be assumed that capabilities that have not been freely available so far will soon be available to a broader mass, and thus also to criminals and actors of opposing states. "This in turn raises questions of national and European security and sovereignty," says Plattner. Her agency has not yet been able to test Mythos, but gained insight into its functionality in a personal conversation with Anthropic developers.
(mho)
