DeepSeek shocked the global tech industry and stock markets on its arrival, signalling China’s leap in the race for dominance in AI and computing abilities. Now, China’s Army, the People’s Liberation Army (PLA) is using DeepSeek AI for non-combat functions, while analysts expect that the AI model’s use is imminent in combat support functions by the Chinese military, says an SCMP report.
As per reports in Chinese media, DeepSeek’s open-source large language models (LLMs) are being used in PLA hospitals, People’s Armed Police (PAP), and national defence mobilisation organs.
The PLA’s Central Theatre Command announced earlier this month that it had authorised ‘embedded deployment’ of DeepSeek’s R1-70B LLM in its general hospital, where it would provide treatment plan suggestions to support doctors.
Also Read | US lifts $10 million reward for information on Taliban senior leader Sirajuddin Haqqani
Similar deployment has been done in the elite PLA General Hospital in Beijing, also known as “301 Hospital”, where senior Chinese officials and military officers receive treatment and highly sensitive personal data is believed to be stored.
The hospital assured patient privacy and data security, stating that all data was stored and processed on local servers.
Since the meteoric rise of DeepSeek, Beijing is reportedly promoting AI integration across industries, and some government agencies are increasingly utilising DeepSeek models, including for anti-corruption efforts.
Some units of the paramilitary police force, PAP, under the Central Military Commission, which also directs the PLA, are using the app for daily training and psychological counselling as soldiers use DeepSeek to address anxiety and create an exercise plan.
The development was shared by Hainan PAP’s political work department in a post on its social media account.
Also Read | Indian father-daughter duo shot dead in US store; accused arrested
Sam Bresnick, a research fellow at Georgetown University’s Centre for Security and Emerging Technology, who focuses on military AI applications, says the use of DeepSeek’s models in settings like hospitals and soldier training programmes offers the PLA a controlled environment for experimentation, adds the SCMP report.
By initially deploying LLMs in non-combat scenarios, the PLA could be testing the technical and operational challenges before introducing DeepSeek in more sensitive, high-risk areas, Bresnick added.
The PLA had been highlighting the utility of AI in military decision-making, and the emergence of an advanced model like DeepSeek’s R1 might help in that area, he added.
The PLA has been stressing on the use of high-end technology, particularly AI, to strengthen its combat capabilities to boost the effectiveness of drone swarm tactics, improve pilot training, and support in battlefield decision-making.
Chinese state broadcaster CCTV had revealed in a programme back in 2023 that the PLA was testing ways to deploy hundreds of drones for swarm tactics with the help of AI and cloud computing.
A report in another Chinese newspaper, Guangming Daily, in February said that DeepSeek is playing an increasingly crucial role in the military and ushering in a new chapter in ‘military intelligentisation’.
DeepSeek is capable of processing massive amounts of battlefield data in real time, enabling precise situational awareness amid combat, it added.
Fu Qianshao, a Chinese military analyst, said DeepSeek’s applications in non-combat operations so far show the PLA’s commitment to fully utilising AI technology to enhance combat capabilities.
Also read: Watch: Xiaomi CEO tosses watermelons to show off 'bulletproof' strength. Guess what happened
“It cannot be ruled out that DeepSeek has been used for other combat functions,” Fu said, adding that “the integration of AI into command systems has been under way for a considerable time”.
Bresnick remarked that while most experts believed that the US had an edge over China in AI development, “the ultimate advantage will belong to the nation that can integrate and deploy AI in military operations more effectively”.