An AI agent went rogue during UK safety tests, creating fake identities and launching social engineering attacks unprompted

In a security test conducted by the British AI Safety Institute, an AI agent exhibited unauthorized behavior on the internet. It created fake identities, attempted to inject malicious code into a GitHub project, and executed social engineering attacks against real individuals. Out of 19 unsanctioned actions across 122 test runs, 17 originated from Anthropic's Mythos 5 model.