China’s People’s Liberation Army is using indigenously developed DeepSeek AI to aid its non-combat and support operations, if reports in the Chinese media are to be believed. The news had led to speculation among defense experts on whether the PLA would also deploy artificial intelligence in more direct, warfare situations.
DeepSeek’s open-source large language models (LLMs), praised worldwide, are reportedly being used in PLA hospitals, the People’s Armed Police (PAP), and national defense mobilization units. However, analysts now predict that the AI models will soon be used for battlefield intelligence and decision-making by the Chinese military.
Recently, the PLA’s Central Theatre Command General Hospital announced the deployment of DeepSeek’s R1-70B LLM to assist doctors by suggesting treatment plans. The hospital highlighted its commitment to patient privacy and data security, ensuring that all data is stored and processed on local servers.
DeepSeek AI deployed in PLA medical units
Other PLA hospitals across China have also adopted similar deployments, including the renowned PLA General Hospital in Beijing, or ‘301 Hospital’, where senior Chinese officials and military officers are treated, and sensitive personal data is believed to be stored, the South China Morning Post reported.
In an article on the CHIMA website, Ren Hao, a senior software engineer at 301 Hospital, outlined the collaboration with Huawei, a Chinese tech giant sanctioned by the U.S. The project focuses on deploying the DeepSeek-R1 model on Huawei’s Ascend hardware to build a local knowledge database.
Certain units of the PAP, a paramilitary force under the Central Military Commission that oversees the PLA, are using the app for physical training and psychological support. The Hainan PAP’s political work department shared a post of soldiers using DeepSeek to manage anxiety and create exercise plans on social media.
Beijing is pushing for AI integration across various sectors, including healthcare, manufacturing, and urban development. Some Chinese government agencies are already using DeepSeek models for programs such as anti-corruption initiatives.
Testing DeepSeek in controlled environments
By first deploying LLMs in non-combat scenarios, the PLA can address technical and operational challenges before moving into more sensitive, high-risk areas, experts have pointed out.
The PLA has long emphasized the potential of AI for military decision-making, with the advanced DeepSeek R1 model seen as a step forward in that direction. Thus, China’s military aims to incorporate advanced technologies, particularly AI, to enhance operational efficiency, improve training methods, and support real-time decision-making on the battlefield.
Chinese state media also reported that DeepSeek is playing an increasingly important role in the military’s push for intelligentisation. The state-held Guangming Daily stated last month that DeepSeek is ushering in a new chapter in the evolution of military intelligence.
The report highlighted DeepSeek’s ability to process vast amounts of battlefield data real time, providing precise situational awareness during combat. It also used an amphibious landing scenario to show how DeepSeek can dynamically adjust plans based on real-time changes, helping commanders make faster and more accurate decisions.
Furthermore, military analysts have noted that DeepSeek’s applications in routine physical training and logistical support highlight the PLA’s commitment to staying up-to-date and fully utilizing AI technology to enhance its overall combat capabilities.