Chinese researchers connected to the People’s Liberation Army (PLA) have created a military AI tool using Meta’s Llama model, which is freely available. According to academic papers, six researchers from three top Chinese institutions, including two linked to the PLA’s Academy of Military Science, used an early version of Llama to build an AI tool called “ChatBIT.” This tool is designed to help with military intelligence by gathering and analyzing information for decision-making.
The researchers improved Llama by adding their own data to create a model focused on military needs. They said ChatBIT was trained for conversations and answering questions related to military situations and claimed it performs nearly as well as OpenAI’s ChatGPT-4.
Meta openly shares its AI models like Llama but restricts their use for military or warfare purposes. However, since these models are public, enforcing these limits can be challenging. Molly Montgomery, Meta’s public policy director, stated that any military use of Llama is against Meta’s policy.
This development is part of a larger trend in China to advance its own AI technology. Reports show that many Chinese researchers, including those from defense-related institutions, are using AI for tasks like electronic warfare and police work.
Source: von
