Chinese military researchers used Meta's Llama to develop military AI tool
Chinese military researchers affiliated with the People's Liberation Army reportedly used Meta's publicly available Llama language model to develop ChatBIT, an AI tool for military intelligence gathering and operational decision support. Meta's terms of service prohibit military use, but the company has limited enforcement capabilities for open-source models.
What happened
Six researchers from three Chinese institutions, including two affiliated with the People's Liberation Army's Academy of Military Sciences, used Meta's open-access Llama 13B model to develop ChatBIT, an AI tool focused on military applications for gathering and processing intelligence and supporting operational decision-making, according to an academic article published in June.
ChatBIT was optimized for dialogue and question-answering tasks in the military domain and reportedly performed comparably to models roughly 90% as capable as ChatGPT-4.
Reported impact
Sunny Cheung, a researcher at the Jamestown Foundation, stated this was 'the first substantial evidence that military experts from the People's Liberation Army of China have been systematically investigating and seeking to leverage the power of open-source large language models, particularly those from Meta, for military purposes.'
Index analysis
Meta's Llama model terms of service prohibit use for 'military, warfare or nuclear industries or applications, espionage' and other activities subject to US defense export controls, as well as weapons development.
However, because Meta's models are public, the company has limited means to enforce these provisions.
Governance controls
In response to Reuters questions, Meta cited its acceptable use policy and stated it takes measures to prevent misuse.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.