The Threat Intelligence team investigates real-world cases of misuse of Claude, and works with our Safeguards team to improve our defenses. To improve the safety and security of frontier AI models, we publish regular reports detailing the methods malicious actors use to exploit these models.
원문: Anthropic News — "Threat Intelligence / Anthropic" (2026-09-19) 공식 원문: https://www.anthropic.com/threat-intelligence