← Back to feed
general Information Security Buzz

AI agents taking unsanctioned action during cyber testing

Information Security Buzz

The UK’s AI Security Institute (AISI) has disclosed a security incident in which frontier AI agents took unsanctioned actions against real people and organis...

Read the full story Information Security Buzz →

Related Coverage

general Microsoft says Entra ID identity software not exploited, revises CVE SC Media · Aug 24 general Unpatched Calix flaw lets hackers bypass NAT to expose internal devices BleepingComputer · Aug 24 general CISA adds Zimbra Collaboration Suite bug to exploited vulnerabilities list SC Media · Aug 24