Washington confronts China’s open-source models

The U.S. government is considering measures to address the use of powerful Chinese open-source AI models, which American labs claim were developed using stolen synthetic data from U.S. models. Strategies under discussion include designating the models as supply-chain risks and tightening export controls, while Nvidia advocates for loosening restrictions to maintain U.S. dependency. Chinese models are being adopted by U.S. startups due to lower costs, despite security concerns.