China probes DeepSeek, Moonshot AI over data security concerns

DeepSeek: DeepSeek is a Hangzhou-based Chinese artificial intelligence company focused on developing advanced large language models with open-weight releases. It has been prioritizing domestic chip usage, including plans for Huawei training chips, while preparing for a potential domestic IPO. In the current news, DeepSeek is one of the primary targets of a probe by China’s internet regulator into data security practices following Anthropic’s accusations of routing sensitive user data through Claude models. Anthropic: Anthropic is a San Francisco-based AI company that develops the Claude family of models with a strong emphasis on safety and alignment research. It recently released Opus 5.5 and issued a detailed threat intelligence report documenting misuse of its models. In the current news, Anthropic’s accusations of illicit data routing by Chinese firms prompted the regulatory probe into DeepSeek and Moonshot AI. Moonshot AI: Moonshot AI is a Beijing-based Chinese AI startup known for its Kimi series of large language models. It has pursued international expansion and IPO preparations on the Hong Kong exchange amid rapid model development. In the current news, Moonshot AI is under investigation by Chinese authorities for data security issues after Anthropic alleged the company routed user requests through its Claude models for distillation purposes. Regulation: China’s Cyberspace Administration has opened an investigation into data security practices at select Chinese AI companies following allegations of unauthorized interactions with foreign models. Industry Practice: AI companies are increasingly publishing detailed reports on model misuse, including distillation attempts by competitors, to highlight risks in cross-border data flows.