The Trump administration is quietly working to ban Chinese open-source AI models from the US. According to Axios, Commerce Department officials considered adding Chinese AI labs to the Entity List, the NSA explored advisories discouraging their use, and draft rules circulated to target Chinese open-source models using supply chain authorities. Pro-innovation voices inside the administration killed those efforts last year. But personnel shifts and the rise of Kimi K3 have given national security hawks new ammunition. David Sacks put it bluntly: "The leading closed labs, already a duopoly in terms of AI model revenue, want the government to eliminate their open-source competition." This isn't really about China. OpenAI and Anthropic have been pushing for licensing and safety regimes that would kneecap open-source competitors for years. Chinese models just give them a national security wrapper to sell it in. The playbook is to use fear of a foreign threat to justify consolidating power domestically. The result would be a US AI ecosystem controlled by two companies with regulatory moats no startup could climb.