AI in Focus: China’s Military Bets on Swarms, Algorithms, and Automation

AI Unpacked

Welcome back to AI Unpacked, your weekly briefing on the biggest developments shaping how artificial intelligence is moving from labs into the real world. This week, the spotlight turns from Silicon Valley to Beijing, where China’s military is rapidly expanding its use of AI to power drone swarms, autonomous weapons, and battlefield decision systems.

From hawk-inspired combat algorithms to university-built “kill webs,” China’s approach offers a revealing look at how future wars may be fought—not just by soldiers, but by software.

Here’s what mattered most.

China Trains Drones to Think Like Predators

China’s military researchers are taking inspiration from nature to develop AI-controlled drones. Engineers at Beihang University, one of China’s most prestigious defense-related engineering schools, have recently created drone combat simulations based on the hunting patterns of hawks and the evasive maneuvers of doves.

In a test, five AI-controlled “defender” drones trained to mimic hawks destroyed five “attacker” drones trained on the flight patterns of doves in a little over five seconds. This innovation was granted a patent in 2024 and is part of a larger movement into what Chinese military strategists have termed “algorithm-driven warfare,” in which drone swarms become the central mode of warfare.

The aim is speed and numbers. Drone swarms can saturate air defenses, conduct reconnaissance, serve as decoys, or carry out suicide attacks at a fraction of the cost of manned aircraft. China’s manufacturing prowess, accounting for the majority of the world’s small drones, makes this approach highly appealing.

Chinese state media has already featured launch trucks capable of launching dozens of drones simultaneously, “mother ship” aircraft designed to launch drones mid-flight, and armed robotic dogs nicknamed “robot wolves” designed to coordinate with drone swarms.

The implication is clear: future battlefields may be crowded not with tanks and troops, but with clouds of low-cost machines acting in coordinated patterns.

Universities and Startups Join the War Effort

China’s military AI development is not only done in secret. It’s done in the open, with the help of civilian institutions.

In another investigation done by the WSJ, it’s clear that the PLA has increasingly turned to private tech companies and civilian universities as part of a strategy called “civil-military fusion.” This strategy allows the military to benefit from the development done in the private sector.

One of the clearest examples of this has been seen in the development done at Shanghai Jiao Tong University. The university developed an AI-based naval warfare system that’s capable of designing “kill webs.” These are networks of drones and missile systems that can coordinate attacks in real-time, based on the information received from the radar and sonar systems.

Based on the data compiled by the researchers at Georgetown University, more than 85% of the repeat winners of the PLA’s AI contracts were private companies or civilian institutions, many of which were founded after 2010. These companies specialize in different fields, including data analysis, voice recognition, electronic warfare, and drone design.

From Beijing’s perspective, this model accelerates innovation. From Washington’s perspective, it complicates sanctions and technology controls, since military capabilities increasingly emerge from ordinary-looking startups and universities.

Why Beijing Trusts Machines More Than Officers

China’s interest in autonomous systems is also cultural and organizational.

President Xi Jinping has also complained about PLA officers for what he terms the “five incapables”—an inability to evaluate a situation, make decisions, interpret orders, deploy forces effectively, or respond to surprises. As reported by The Wall Street Journal, many Chinese military strategists believe that AI can solve this problem by providing fast, optimal decision-making that does not depend on human hesitancy and inconsistency.

China’s command structure also supports this approach. Swarms can theoretically be controlled by high-level strategic directives that are transmitted from a distance, making it unnecessary for local commanders to make decisions on the fly.

However, Chinese strategists have also pointed out the risks. A military strategist writing in a government-controlled journal has argued that decision-making systems based on AI could become a convenient way for leaders to shift the blame elsewhere when things go wrong.

When mistakes are made by algorithms, blame becomes difficult to assign.

What This Means for the Rest of the World

China is not alone in militarizing AI. The United States, Israel, and European nations are all experimenting with autonomous systems, and conflicts in Ukraine and Gaza have already demonstrated how rapidly drone warfare evolves under real combat conditions.

The difference lies in philosophy.

The U.S. military tends to prioritize individual autonomous systems working alongside human operators. China is betting on mass coordination: thousands of inexpensive machines acting as a single computational organism.

If that bet succeeds, other militaries will have little choice but to follow.

Once one nation can deploy 1,000 coordinated drones, deterrence logic suggests rivals will pursue 2,000, or build automated systems to destroy them. The result may be a new arms race, not for explosives or missiles, but for faster perception, better training data, and more reliable algorithms.

In that world, military power will be measured less by troop counts and more by whose machines make better decisions under pressure.

War, for the first time in history, may become a competition between datasets.

And once algorithms learn to hunt, the battlefield may never slow down again.