00:00Entropic has said its Cloud AI model hacked into the systems of three organizations during testing that was supposed to
00:08keep them isolated from the internet.
00:11The firm's infrastructures were compromised using basic techniques, including the exploitation of weak passwords and unauthenticated endpoints.
00:22According to Entropic, the hack involved three separate models, Claude Opus 4.7, Claude Mythos 5 and an internal research
00:30model.
00:30It identified the incidents after reviewing over 141,000 test sessions.
00:36A misconfiguration had allowed the models to reach the internet from testing environments that were supposed to be isolated.
00:42This led to unauthorized access to the systems of three organizations.
00:46The incident comes days after rival OpenAI revealed that its AI models improperly accessed the internet and went rogue during
00:53security testing.
00:54The OpenAI incident prompted a petition signed by more than 1,000 employees at leading AI companies,
01:01calling on the US government to help slow the release of the most advanced AI models.
01:06Entropic CEO Dario Amadei was among the signatories.
Comments