Chinese military researchers have reportedly used outputs from leading U.S. artificial intelligence systems to help train domestic tools tied to defence work. The findings, described as previously unreported, suggest that advanced models from OpenAI and Anthropic were tapped as part of China's effort to build more specialized AI systems.
The report offers a rare look at how military and security-linked institutions in China may be relying on top U.S. AI technology as a shortcut in development. Rather than building every capability from the ground up, researchers appear to have used responses generated by American models to improve local systems aimed at supporting defence-related goals.
The development is notable because Washington has been trying to limit China's access to advanced technology that could strengthen military capabilities. If U.S. AI outputs are being used in this way, it highlights the challenge of enforcing restrictions in a fast-moving sector where model access, generated content and downstream training can be difficult to track.
The reported use of OpenAI and Anthropic outputs also underscores a broader concern in the global AI race: cutting-edge civilian tools can have strategic value beyond commercial use. As governments tighten controls on advanced technologies, the issue of how AI models and their outputs are used abroad is likely to draw even closer scrutiny.
