China Adapts Meta’s Llama AI Model for Military Use

Researchers affiliated with China’s People’s Liberation Army (PLA) have reportedly adapted Meta’s open-source Llama AI model to create a military-focused tool known as “ChatBIT.” According to academic papers reviewed by Reuters, Chinese research teams used an early version of Meta’s Llama model to enhance military decision-making and intelligence gathering.

Development of ChatBIT and Potential Military Applications

Chinese researchers, including members from the PLA’s Academy of Military Science, detailed how they customized Meta’s Llama 13B language model for ChatBIT, an AI tool optimized for military use. According to the researchers, ChatBIT supports tasks such as intelligence analysis and operational planning, aiming to provide accurate information for strategic decision-making. While ChatBIT has been fine-tuned for dialogue tasks in the military field, its full capabilities and deployment status remain unknown. However, it was found to outperform certain AI models that are approximately 90% as capable as OpenAI's advanced ChatGPT-4.

"It's the first time there has been substantial evidence that PLA military experts in China have been systematically researching and trying to leverage the power of open-source LLMs, especially those of Meta, for military purposes," said Sunny Cheung, an associate fellow at the Jamestown Foundation.

Research Team and Future Applications

The Chinese research team behind ChatBIT includes Geng Guotong and Li Weiwei from the Academy of Military Science’s (AMS) Military Science Information Research Center and the National Innovation Institute of Defense Technology, along with researchers from the Beijing Institute of Technology and Minzu University. According to the research paper, “In the future, through technological refinement, ChatBIT will not only be applied to intelligence analysis, but also ... strategic planning, simulation training and command decision-making will be explored.”

Meta’s Response and Security Concerns

Meta’s terms prohibit the use of its AI models for military purposes, but as Llama is open-source, enforcing these restrictions remains challenging. Meta responded to the reports by affirming its stance against unauthorized use, with Public Policy Director Molly Montgomery stating, “Any use of our models by the People's Liberation Army is unauthorized and contrary to our acceptable use policy.”

Meta argued that open-source models are vital to AI innovation but acknowledged concerns around potential misuse. "In the global competition on AI, the alleged role of a single, and outdated, version of an American open-source model is irrelevant when we know China is already investing more than a trillion dollars to surpass the US on AI," a Meta spokesperson noted.

U.S. Monitoring and Regulatory Actions

Amid rising tensions over AI’s role in defense, the Pentagon has confirmed it is closely monitoring advancements in competitors’ AI capabilities. In October 2023, President Biden signed an executive order to address AI security risks, and Washington has since taken steps to regulate U.S. investments in China’s AI and tech sectors.

Pentagon spokesperson John Supple emphasized the dual nature of open-source AI: “We will continue to closely monitor and assess competitors' capabilities.”

Concerns Over China’s Growing AI Capabilities

China’s rapid AI advancements, including its adaptation of Western-developed technology, are fueling concerns among U.S. officials and analysts. Researchers note that China’s extensive network of AI research labs and collaborations with international scientists make it difficult to contain its AI development. According to William Hannas of Georgetown University’s CSET, “There is too much collaboration going on between China's best scientists and the U.S.' best AI scientists for them to be excluded from developments.”

Looking Ahead

As open-source AI models continue to shape global tech landscapes, the ethical and security implications of unrestricted access are under scrutiny. The adaptation of Meta’s Llama model for military applications in China highlights the risks and challenges of AI’s role in defense. With AI’s potential to alter defense capabilities, balancing innovation with security remains a pressing issue for governments and tech companies alike.