AnthropicResearch
Anthropic's Anvil automated red-teaming system finds stronger Claude safeguards jailbreaks than human experts
Anthropic published Anvil on November 12, 2026, its automated red-teaming system for Claude safeguards: Anvil beat human red-teamers across bioweapons, cybersecurity and distillation safeguards, finding stronger jailbreaks in 24 to 48 hours instead of weeks.
Image: Anthropic







