Embargoed_Anthropic_full_threat_intel_report.pdf (1 MB)

Anthropic reports that North Korean operatives misused Claude to support fraudulent remote-employment schemes targeting US Fortune 500 technology companies. The activity involved creating convincing false identities, completing technical and coding assessments, and performing real technical work after hiring, with the goal of generating revenue for the North Korean regime despite sanctions. The source frames this as an evolution of an established DPRK IT-worker operation, where AI reduces previous bottlenecks in coding ability, English communication, and specialized training. Anthropic says it banned the related accounts, improved indicator collection and correlation for this scam type, and shared findings with relevant authorities.