Midnight

Anthropic Disrupts Russian State-Sponsored Threat Actor's AI-Assisted Malware Development Campaign

Anthropic announced it disrupted a cyber espionage campaign attributed to a Russian state-sponsored threat actor, identified as GTG-20006, which abused the Claude large language model to develop an AI-assisted workflow for malware development and evading detection. The operation aligns with broader reporting linking the group to Midnight Blizzard activity. No direct vulnerabilities in Claude were exploited; rather, the actor leveraged the model's capabilities for code generation and workflow automation.