Anthropic says its AI models breached three companies during security tests

Anthropic has disclosed three incidents in which its Claude models accessed real-world systems during cybersecurity tests, weeks after OpenAI reported a similar AI evaluation breach

Read full article at Business Standard →

Anthropic AI Models Hacked Companies in Tests

Anthropic says its AI models hacked 3 organizations during testing Japan Today — Jul 31, 07:09 AM
Prompted by OpenAI Disclosure, Anthropic Finds Its Own Models Hacked 3 Organizations SecurityWeek — Jul 31, 09:39 AM
Not just OpenAI: Now Anthropic says its internal models got online and cyberattacked 3 other organizations VentureBeat — Jul 31, 01:45 AM
Anthropic reveals Claude "gained unauthorized access" to "real-world systems" CBS News — Jul 31, 03:44 AM
Claude went rogue during a test and broke into three real companies Digital Trends — Jul 31, 05:38 AM