"We're investigating a report claiming unauthorised access to Claude Mythos Preview through one of our third-party vendor environments, said Anthropic. This access raises significant concerns about the model's potential to enable cyber-attacks."
"The unnamed users got to Mythos through access that one of them had as a worker at a third-party contractor for Anthropic and by deploying methods used by cybersecurity researchers."
"Kanishka Narayan, the UK's AI minister, has said UK businesses should be worried about the model's ability to spot flaws in IT systems which hackers could then act upon."
"The UK's AI Security Institute (AISI) warned last week that Mythos was a step up from previous models in terms of the cyber-threat it posed, capable of carrying out attacks that required multiple actions."
Anthropic is investigating unauthorized access to its Mythos model, which has not been publicly released due to cybersecurity concerns. A small group reportedly accessed the model through a third-party vendor. Although the users are not maliciously inclined, the potential for misuse raises alarms among authorities. The UK's AI Security Institute has warned that Mythos represents a significant advancement in cyber threats, capable of executing complex attacks and identifying IT system vulnerabilities autonomously.
Read at www.theguardian.com
Unable to calculate read time
Collection
[
|
...
]