DeepSeek and Moonshot Allegedly Sent User Requests to Anthropic's Claude Model

Chinese AI companies DeepSeek and Moonshot are alleged to have transmitted user requests to Anthropic's artificial intelligence model Claude, according to a report released by Anthropic on October 10. The company stated that Moonshot forwarded customer inquiries to Claude and then presented the responses as if they were generated by its own model. Anthropic also suggested that some of these requests may have contained sensitive information related to Chinese companies and public institutions.
Anthropic's threat intelligence report further revealed that DeepSeek was found to have similarly sent user requests to Claude and used the responses for training their models. While AP News described this issue as one of the points of contention in the growing AI technology rivalry between the U.S. and China, there is no confirmed investigation by Chinese authorities into DeepSeek or Moonshot.
Korean Source
This article is an English localization of a Korean-language crypto news report. Original headline: 딥시크·문샷, 앤트로픽에 사용자 요청 전송 의혹