Chinese military uses US AI models

A Reuters review of more than 80 Chinese academic papers and patents, including research compiled by the Jamestown Foundation, found that Chinese military researchers linked to the National Liberation Army have used outputs from leading U.S. AI models developed by OpenAI and Anthropic to train domestic AI systems for defense and security applications. The technique, known as model distillation, allows smaller models to be trained on outputs from powerful frontier models, reducing computing costs and enabling local deployment. Applications include processing sensitive military code, drone navigation, target recognition, and social media monitoring. The findings highlight a flashpoint in U.S.-China relations, with U.S. officials accusing Chinese entities of unauthorized extraction that undermines export controls and infringes intellectual property. China has rejected the accusations, accusing Washington of AI 'hegemonism.' The dispute has also involved Moonshot AI, which denied allegations that its Kimi K3 model was built using distillation. Experts note that distilled models have limitations and cannot fully replicate frontier AI capabilities.