Anthropic says safeguards cut theft of Claude reasoning tokens by up to 99%
🤖 AI-generated content — The title and summary were produced automatically by artificial intelligence, without human editorial review.
Anthropic says new safeguards have reduced extraction of Claude’s internal reasoning tokens by 90–99% and led to tens of thousands of account bans. It alleges that Zhipu and Moonshot used large networks of fake accounts in model-distillation attacks, while acknowledging that some legitimate users are also being blocked.
Continue in the app — vote & join in ➔Source: The Decoder (AI) · via ahirlevel.hu