Moonshot AI announced the release of its new Kimi K3 model this week, sparking renewed debate about China's role in open source AI. The company stated that while Kimi K3 still lags behind the most powerful proprietary models like Claude Fable 5 and GPT 5.6 Sol, it demonstrated frontier-level performance across its evaluation suite, outperforming other tested models. Independent analyses from Arena.ai and Vals AI also confirmed Kimi's competitiveness with flagship frontier models. The announcement came as Chinese President Xi Jinping spoke at the World AI Conference in Shanghai, coinciding with a drop in the Nasdaq by about 1% as investors sold off stocks in chip companies like Nvidia. This has raised concerns among tech industry figures, with some drawing parallels to the earlier debate sparked by DeepSeek's release of its open source R1 model in January 2025. However, the current context is marked by heightened tensions following the Trump administration's tariff war with China and ongoing disputes over the national security risks posed by AI models like Anthropic's Claude. The situation has also intensified as major AI companies prepare for public listings.
David Sacks, former Trump administration AI czar and now co-chair of the President’s Council of Advisors on Science and Technology, criticized the U.S. for creating regulatory barriers that could hinder its AI progress. He argued that the U.S. is 'tying itself in knots' with data center bans, state regulations, and federal pre-approvals for frontier models, which could lead to losing the AI race. Sacks also took a shot at Anthropic, calling Claude an example of 'woke lobotomized models.' Former Uber CEO Travis Kalanick echoed these concerns, suggesting that Chinese companies are 'distilling off' American AI models. He warned that if distillation isn't restricted, it would put American models at a disadvantage. However, he acknowledged that American models have also been built on top of Chinese models like Kimi.
OpenAI’s head of strategic futures, Dean Ball, praised Kimi as 'a very good model' whose performance can't be explained by distillation. He expressed surprise that the Chinese state continues to allow the open sourcing of such models, given potential risks. Ball also speculated that a world dominated by open-weight models could lead to 'full AI communism,' where AI is treated as a public good provided by the state. He suggested that the Trump administration might eventually recognize the need to create regulatory risks around the use of open-weight Chinese models. However, he argued that banning open source isn't necessary; instead, creating enough regulatory uncertainty could deter enterprises from using these models.
Source: techcrunch
