Arcee: Chinese AI Models Are No More Dangerous Than Any Other Open-Source Software
Moonshot AI
Alibaba/Qwen
OpenAI
Anthropic
The American open AI lab Arcee argues that Chinese open-weight models such as Kimi K3 and Qwen do not pose a unique threat. Arcee CTO Lucas Atkins explains that these models, running in an isolated user environment, have no communication channels with developers, and their code is available for inspection. Instead of bans, he proposes strengthening the open ecosystem in the United States.
Amid the growing capabilities and popularity of Chinese open-weight AI models (such as Moonshot AI's Kimi K3 and Alibaba's Qwen), debates have reignited in the U.S. over whether to restrict them. The Trump administration is considering a ban on these models, though it has not taken action yet. Proprietary developers, especially OpenAI and Anthropic, have expressed concerns. However, Lucas Atkins, technical director of the American lab Arcee, argues that the fears are overblown. According to him, Chinese open-weight models are no more dangerous than any other open-source software, since once downloaded into a user's environment, the developer loses access to them. The models' source code, available on platforms like Hugging Face, is open for inspection, and large organizations can conduct their own security testing. Atkins notes that while it is theoretically possible to embed malicious code in a model, it is extremely difficult in practice. He believes that instead of banning Chinese models, the U.S. should develop its own open ecosystem. Arcee itself uses Chinese models for training and improving its own developments. In Atkins' view, the best way to compete is to create better open models.
Source: TechCrunch AI —
original