Anthropic details new defenses that cut distillation attacks on Claude
Original titleStrengthening our defenses against distillation attacks
Anthropic says its new defenses have cut the reasoning tokens distillers can extract from Claude by a factor of 10 to 100 on its own platform.
The company describes three measures: controlling where reasoning can go, a two-stage classifier that blocks extraction attempts with a false positive rate below 0.1%, and stricter limits and bans for repeated attempts.
Anthropic also names Zhipu's GLM-5.3 and Moonshot's Kimi as cases where distilled capabilities raised cyber and national security concerns.
The post names specific distillation methods and their measured effects on reasoning extraction, which helps readers judge how AI labs protect models from copying.
Source: Anthropic Research · anthropic.comPublished · added here