Anthropic publishes September threat-intelligence report

On 10 September 2026 Anthropic published a 154-page threat-intelligence report on malicious uses of Claude it stopped between December 2025 and August 2026, across cyber operations, surveillance, fraud, weapons work, and illicit model distillation.

Why it mattered It showed AI moving from assisting cyberattacks to orchestrating them, with multi-agent frameworks running large parts of intrusions, and became the reference account of AI-enabled misuse at scale.