A coalition of American AI startup founders is urging the Trump administration not to ban Chinese open-weight AI models, warning that such a restriction would backfire on U.S. innovation, according to Politico.
The founders argue that cutting off access to open-weight models — including those released by Chinese labs — would force smaller American companies to rely on expensive, proprietary systems from incumbents like OpenAI and Anthropic, effectively creating a duopoly among large closed-source AI firms. Open-weight models make their underlying parameters publicly available, letting startups and researchers build on existing work rather than training models from scratch, which lowers costs and speeds development.
Washington policymakers have raised the possibility of restricting or banning models from China over national security concerns, including fears that such systems could be repurposed for cyberattacks, disinformation, or weapons development. The founders counter that the stronger safeguard against those risks is a competitive, fast-moving domestic AI industry, and that restricting access to widely used open-weight tools would slow that industry down rather than protect it.
The dispute sets up a broader policy fight over whether the U.S. should treat open-weight AI as a national security liability or as an innovation resource — one Politico reports is now playing out directly with the administration.