Anthropic draws a red line: The developer of the Claude chatbot will in future prohibit business with companies that are majority-owned by Chinese hands.

Not only tech giants like ByteDance, Alibaba, or Tencent are affected.

The step comes at a delicate time. In Washington, there is growing concern that China is acquiring US technology through shell companies abroad. At the same time, the example of the start-up DeepSeek, which delivered competition at eye level with its open R1 model, shows how quickly Chinese players are catching up. OpenAI had previously published evidence that DeepSeek had illegally used US models for training.

For Anthropic, the new policy means revenue losses in the "low hundreds of millions". However, the founders, who established the company in 2021 out of safety concerns about uncontrolled AI development from OpenAI, see it as a political signal: From the company's perspective, AI innovations should serve democratic interests and not fall into the hands of geopolitical rivals.

The decision comes immediately after a new financing round of 13 billion dollars, valuing Anthropic at 170 billion dollars. While the Biden administration has already tightened export controls, Anthropic is now setting its own regulatory accents – thereby accepting that competitors may benefit in the short term.