Russia-linked hackers used Anthropic's Claude AI for espionage, disinformation, drone ops

Anthropic's report reveals Russian-linked hackers used its Claude AI model to automate attacks on over 20 targets, including Ukraine's Ministry of Defense and drone suppliers, and to run pro-Kremlin disinformation campaigns in Africa and Moldova. This marks a notable escalation in AI-enabled cyber operations by state-linked actors, raising concerns about dual-use AI models. The full scope of impact and attribution details remain partially redacted.