Trump Administration Weighs Options as Chinese AI Models Advance
The Trump administration is divided over how to address the rapid progress of Chinese artificial intelligence models, according to people familiar with internal discussions.
The dispute reportedly centers on differing approaches within the White House and Commerce Department. Some White House officials favor tighter restrictions designed to prevent Chinese laboratories from using leading American models to improve their own systems. The Commerce Department, which oversees export controls through the Bureau of Industry and Security, has questioned whether such measures would be practical or enforceable.
The debate intensified after Moonshot AI released Kimi K3, a model described as competitive with advanced systems from Anthropic and OpenAI. Administration officials are considering steps aimed at limiting “distillation,” a process in which one model is trained using outputs from another. Anthropic recently alleged that Alibaba conducted a major distillation effort targeting its technology. The White House is expected to respond through presidential action, though not necessarily an executive order.
Officials have publicly characterized covert distillation as intellectual property theft and raised the possibility of sanctions. However, no formal White House proposal has reportedly been sent to Commerce.
Commerce Secretary Howard Lutnick has also explored incentives for American developers to release open-weight models that could compete with Chinese offerings. This approach would seek to strengthen US alternatives rather than rely solely on restrictions.
The policy discussion reflects broader concerns about open-weight AI systems, which can be downloaded, modified, and deployed with limited safeguards. Administration officials are also assessing the security implications of powerful American models capable of identifying vulnerabilities in government networks and critical infrastructure. The outcome could shape how the United States balances competition, national security, intellectual property protection, and model accessibility. No final policy has been announced, and the internal debate remains unresolved.