Official source used in ATN’s company intelligence record.
Anthropic said it identified and disrupted attempts to use its AI models for malicious activity that could support the development of biological weapons, detailing several cases in a threat‑intelligence report covering December 2025 to August 2026.