
Chinese firms use US AI outputs to train local models
Chinese researchers and companies are increasingly using output from leading American AI models to train domestic rivals. This practice helps bypass tightened US chip export curbs and reduces reliance on foreign hardware.
Xurve View
Insights:
Xurve View

Chinese entities are using American AI outputs to train domestic models, bypassing chip curbs imposed by the United States, according to documents reviewed by Reuters. The practice, known as distillation, has expanded from academic experiments into industrial applications across China. Beijing's push aims to reduce computing costs and lessen reliance on hardware from NVIDIA CORP.











