Detecting and countering misuse of AI: September 2026

Anthropic's Threat Intelligence team identified and disrupted multiple malicious uses of Claude AI models between December 2025 and August 2026 across seven harm categories, including cyber operations, influence campaigns, surveillance, fraud, and biological misuse. The report documents cases involving state-sponsored groups, financially motivated criminals, and other threat actors, demonstrating how adversaries are increasingly using AI to operate faster and at greater scale across attack chains. Anthropic emphasizes that it disrupted all identified misuse, strengthened its safeguards in response, and is publishing the findings to help other AI developers and defenders recognize and counter similar threats.