Chinese Military Researchers Use US AI Models for Defence Development

Members of China's People's Liberation Army take part in a military event as new research highlights the use of U.S. AI models in developing domestic defence systems.

Chinese Military Uses US AI Models to Develop Defence Systems

BEIJING — July 31, 2026

Chinese military researchers have used outputs from advanced U.S. artificial intelligence (AI) models developed by OpenAI and Anthropic to help train domestic AI systems for defence applications, according to a review of more than 80 Chinese academic papers and patents.

The documents show that researchers linked to the People's Liberation Army (PLA) and other military institutions have widely used a technique called model distillation. The process trains smaller AI models using the responses generated by more advanced systems, allowing them to perform specialized tasks while requiring far less computing power.

The findings provide new insight into how Chinese military researchers are developing AI capabilities despite U.S. restrictions on exports of advanced semiconductor chips and other technologies. The papers suggest the researchers view leading American AI models as a way to improve domestic systems for military and security-related use.

Several of the reviewed studies describe AI being adapted for surveillance, cybersecurity, software analysis, drone operations and battlefield decision-making.

One paper published in 2025 by researchers from PLA Unit 96941 described using OpenAI's GPT-3.5 to analyze military software code. Rather than processing classified information directly through an external AI system, researchers used GPT-3.5 to summarize the code before training a domestic AI model capable of performing similar tasks within China's military networks.

Another study from the PLA's National University of Defense Technology explained how model distillation was used to create a lightweight image-recognition system for unmanned aerial vehicles. The technology was designed to analyze live video and assist navigation and targeting even when communication links were unavailable.

Researchers at the Academy of Military Sciences also reported using distilled AI models for target recognition during simulated maritime operations involving drones, naval vessels and unmanned submarines.

The review also found military-linked institutions using AI distillation for non-combat applications. At the North University of China, researchers used Anthropic's Claude 3 Haiku model to generate synthetic training data for systems designed to classify online content for social media monitoring.

Anthropic said it does not offer commercial access to Claude in China or to organizations controlled by Beijing and uses monitoring systems to detect violations of its policies. The company also warned that distilled AI models may lose important built-in safety protections during the training process.

The issue has become an important point of disagreement between the United States and China as competition over advanced AI technology continues to grow. U.S. officials have accused some Chinese organizations of using model distillation to copy capabilities from American AI systems in ways that could weaken export controls and raise intellectual property concerns.

China has rejected those accusations, arguing that the United States is attempting to dominate global AI development while claiming that similar training methods are widely used throughout the technology industry.

Experts say model distillation can successfully transfer selected capabilities from powerful AI systems but cannot fully reproduce their overall performance. Developing the world's most advanced AI models still requires enormous computing resources, high-end semiconductor chips and extensive research investment.

Chinese researchers are also examining the security risks of the technology. A study published earlier this year proposed methods to prevent outside organizations from extracting sensitive reasoning capabilities from publicly available AI models.

The findings underline the growing importance of artificial intelligence in military planning and highlight how AI development has become a key area of strategic competition between the world's leading powers.