The drones that fanned out during a recent People’s Liberation Army exercise were dispatched by the Intelligent Precision Strike System, a new product from Chinese defense giant Norinco that used the UAVs’ real-time data to model the battlefield, track targets, devise strike plans, distribute firing information, and execute follow-up strikes.
According to the video playing in Norinco’s booth at the most recent Zhuhai Air Show, almost all of this was done autonomously except giving the commands to fire. Chinese observers also noted how the system fused battlefield intelligence from multiple sources. It epitomizes how the PLA aims to ensure dominance in the next era of conflict: with autonomous capabilities that blur the line between human oversight and machine execution.
Norinco’s Intelligent Precision Strike System is one of the ways that the nascent PLA Information Support Force is building a “network information system” that uses AI, cloud computing, and big-data techniques to fuse data from operational units and create “dynamic kill networks” across domains. PLA commentators emphasize the network information system’s critical role in modern warfare.
All this is part of the PLA’s push towards “intelligentized warfare,” which seeks to integrate real-time battlefield awareness, precision strikes, and psychological operations. To this end, researchers at the PLA National Defense University highlight large language models as pivotal to military operations. By quickly processing vast datasets, they could streamline intelligence analysis, generate code, and accelerate weapons development.
The PLA also aims to use LLMs to produce detailed, realistic operational simulations and training scenarios with a fraction of the time and manpower required today. Taiwan’s Institute for National Defense and Security Research has called attention to the PLA’s “War Skull” wargaming system, whose second generation, launched in 2020, uses modular strategies to adapt to various adversaries—although its applicability to real-world combat remains uncertain due to challenges like emergent behaviors and unpredictability in complex scenarios.
Similarly, China is seeking to embed AI in military intelligence, planning, and decision-making. The “Aiwu LLM+” system, developed by a lab at the People’s Armed Police Engineering University, combines large language models, multimodal big-data analysis, and virtual assistant interfaces to provide intelligent interaction and task planning within command information systems. The PLA also envisions the integration of AI into multi-source intelligence systems that provide commanders with insights and accelerate decision-making.
All this is viewed as helping the PLA move into the next phase of intelligentized warfare, where deep learning and multimodal data processing can refine target recognition, situational assessment, and command decisions. According to PLA commenters, these advancements facilitate iterative feedback loops, improving cross-domain data integration, predictive analytics, and real-time battlefield adaptation—and ultimately forming what the PLA calls “intelligentized operational command.” Meanwhile, integrating “shallow AI” into unmanned platforms bolsters reconnaissance and precision strikes while embedding AI-driven autonomy into existing weapons.
The gains are not just about kinetic effects. PLA Maj. Gen. Zeng Haiqing notes that LLMs could also be used for cognitive warfare, which PLA doctrine describes as the key to winning wars, including by not even having to fight them. Cognitive domain operations merge psychological and cyber tactics, seeking to manipulate adversary perceptions, decision-making, and behavior. Generative AI tools enable the PLA to craft adaptive, context-specific disinformation and execute psychological operations with precision. Advanced language models can generate desired narratives in real-time, using digital platforms to influence perceptions, sow discord, and erode morale. This serves the PLA’s “cognitive confrontation” tactic, which aims to control information flows to disrupt adversary decision-making. AI-powered sentiment analysis and predictive behavioral models can maximize these strategies’ psychological and operational impact.
These AI initiatives are accompanied by the PLA’s growing integration of big-data programs that can expand battlefield awareness, refine predictive analytics, and reduce the “fog of war.” Intelligent algorithms can process vast datasets to uncover operational patterns, optimize logistics, and better tactical decision-making.
Industry help
Under China’s military-civil fusion strategy, defense contractors large and small are working to ensure that civilian AI advances can power military applications.
Norinco, for example, showed off more than the Intelligent Precision Strike System at the 2024 Zhuhai Airshow. Nine other new combat systems with cutting-edge AI-enabled warfare capabilities were on display. They included the AI-Enabled Synthetic Brigade, which combines next-generation armored vehicles, swarming drones, loitering munitions, and electronic warfare tools; and the Smart Digital-Enabled Command and Control System, which enables real-time situational awareness. Their effect has already been felt in human confidence levels. Chinese military analysts claim these AI-driven mechanized brigades surpass their U.S. counterparts in battlefield digitization, reinforcing China’s leadership in next-generation land warfare.
Smaller tech firms are also contributing. For example, U-Tenet has developed military-focused AI models and systems that support strategic decision-making and autonomous operations. These include Tianji, a cloud-based “decision-making brain” for operational planning and intelligence analysis; Tianwang, a real-time intelligence repository that integrates multi-source data for situational awareness; and Tianjian, an integrated battlefield-intelligence system. U-Tenet built its AI applications with a proprietary military-intelligence database that contains more than a million high-quality documents and more than 300 terabytes of military imagery, according to ifenxi, a Chinese digital research and consulting firm. The Tianji model can incorporate real-time conflict data, including from the war in Ukraine, Chinese military commentators report.
The success of the PLA’s hefty investment in AI-infused tools will hinge on its ability to validate and refine these technologies under real-world conditions. Their challenges include ensuring that they work reliably in complex scenarios and integrating them into a centralized control structure. The PLA’s goal with AI and Big Data, however, is clear: not just to close critical capability gaps but to redefine what is possible in warfare.
