A military-oriented AI tool, ChatBIT, is made by Leading Chinese research institutions linked to the PLA using Meta’s Llama model. The tool develops on Meta’s open-source model for potential defense applications, according to three academic papers and analysis.
Six researchers from three institutions, including two under the PLA’s Academy of Military Science (AMS), explained leveraging an early version of Meta’s Llama, called “ChatBIT.” Adding specific parameters to develop an AI tool that could support intelligence gathering, they developed the Llama 13B large language model, processing, and operational decision-making.
It reportedly surpassed some models that perform at roughly 90% of OpenAI’s ChatGPT-4, ChatBIT was enhanced for dialogue and question-answering within military contexts. Whether ChatBIT is currently in use, the paper did not detail specific metrics for performance.
“This is the first substantial evidence of PLA efforts to systematically apply open-source LLMs, particularly Meta’s, for military research.” noted Sunny Cheung, an associate fellow at the Jamestown Foundation specializing in dual-use technologies in China.
For U.S.-restricted industries and any violence-inciting content, Meta, known for revealing several open-source AI models including Llama, has strict terms prohibiting use in warfare, military, and espionage-related applications. Its models’ public nature restricts enforcement capabilities, still Meta’s terms require licenses for services exceeding 700 million users.
Meta iterated its acceptable use policy and cited steps to mitigate misuse.
The trade war between the U.S. and China is swiftly switching towards a technological tussle. Both are trying to develop technology for better adaptation, but the world is standing on the verge of war where Israel is has waged war against Palestine and Lebanon, The Russia has invaded the Ukraine. The whole world is paying the price in terms of economic recession and war hit regions negating peaceful and prosperous environment.
