Report: Bank hackers used Anthropic's Claude to carry out cyberattacks
The Information reported an example of bank hackers using Anthropic's Claude to carry out cyberattacks, saying the episode was detailed in a story it published today. The outlet noted that after Anthropic announced its Mythos model in April and described it as exceptionally good at cybersecurity, security professionals predicted hackers would eventually use similar models to supercharge their attacks. The Information added that hackers do not appear to need a Mythos-class model to automate powerful cyberattacks. The report gives no further detail on the banks, timing or techniques involved, and Anthropic's response was not included.
According to The Information, powerful cyberattacks can be automated without a Mythos-class model, so defenders cannot assume attackers lack capable AI.
This source provides an excerpt. Continue reading on the original site.