China’s People’s Liberation Army has spent the last few years turning swarm rhetoric into demonstrable capability. State-funded documentaries and PLA media have showcased swarms that can reconfigure after interference, recover lost links autonomously, and shepherd loitering munitions to targets—a capability set that, if operationalized at scale, would reshape air defence calculus in any contested theater.
But public reporting of the PLA’s own exercises suggests that the transition from impressive lab demos to reliable battlefield effects is not yet complete. In a CCTV-backed training vignette reported by the South China Morning Post, a PLA anti-aircraft artillery unit achieved roughly a 40 percent damage rate in a first engagement against a small drone swarm, with officers admitting shooting at swarms remains challenging given target size, speed and maneuverability. That candid result is striking because it comes from the force trying to field the swarm capability.
Why that matters is not merely academic. The last three years of combat in Ukraine and other conflicts have taught militaries a blunt lesson: low-cost loitering munitions and massed small UAS can convert quantity into operational effects. Adversaries can accept attrition math that would be intolerable against high-value platforms, and they can iterate rapidly on production, tactics and deception. That battlefield precedent is one reason the PLA and its industrial partners have poured resources into resilient swarm architectures and distributed command nodes.
Yet those same precedents also reveal practical options for defenders. Think tanks and operational studies argue for layered active defences and complementary passive measures rather than a single silver bullet. Layering means combining electronic warfare, rapid-fire kinetic systems, interceptors, directed energy and passive hardening and deception to increase cost-imposition for an attacker and to create multiple failure points for swarms. Importantly, mobility and integration matter: fixed point defenses can be saturated, while mobile, networked C-UAS nodes provide resilience for maneuver forces.
Recent industry and national tests show how some of those layers are maturing. Western firms and militaries demonstrated mobile laser systems and integrated kinetic plus non-kinetic suites that can reliably defeat small UAS in controlled trials. For example, a coalition integration of a 26 kW LOCUST laser with a Stryker-based C-UAS platform reported highly successful live-fire engagements against Group 1-3 UAS during 2024 demonstrations, illustrating that directed energy is moving from concept to practical edge-defence tool when matched with appropriate sensors and battle networks. Those demonstrations do not mean lasers are a turnkey answer to massed swarms, but they do show an important element of a layered approach finally achieving fieldable maturity.
Pulling the PLA test and Western countermeasures together yields five pragmatic lessons for planners who must confront the prospect of Chinese-style swarming at scale.
1) Build the layer, not the single shot. No single weapon will be cost-effective against thousands of cheap drones. Combine EW, small-kilowatt lasers, high-rate-of-fire guns with airburst munitions, interceptor UAVs and decoys so an attacker must penetrate multiple types of mitigation to reach valuable targets. Recent policy work emphasizes that this layered approach is the most realistic path to operational resilience.
2) Prioritize sensor fusion and AI-enabled battle management over raw kill-tech headlines. The PLA’s own demonstrations stress autonomous swarm recovery and onboard coordination. If defensive systems cannot find or classify hundreds of small, low-RCS targets in cluttered littoral or urban environments, even expensive effectors will be misapplied. Investing in multispectral sensors, rapid target correlation, and AI that supports human-in-the-loop decision making will buy defenders time and efficiency. The PLA admission about poor hit rates shows sensors and engagement chains are the true bottleneck.
3) Field mobile, distributable C-UAS for maneuver forces. Static point defences invite saturation. Mobile directed-energy and kinetic C-UAS nodes that can ride with formations, tether into joint networks, and accept short logistic tails will complicate an attacker’s targeting math. Live demonstrations of vehicle-mounted lasers and integrated systems show this is achievable in the near term when procurement and doctrine align.
4) Invest in cost-imposition economics. Adversaries will exploit cheap production. Defenders must develop low-cost interceptors, large magazines of short-range munitions, and non-kinetic effects like high-power microwave or scalable jamming that change the cost curve for attackers. Lethality alone does not solve the problem; tempo, stockpile depth and sustainment do. Think pieces on counter-swarm strategy argue for stockpiles and high-volume cheap solutions as essential complements to high-tech options.
5) Prepare doctrine, training and legal guardrails for autonomy and escalation. Swarms push commanders toward delegating rapid engagement decisions to machines. That is operationally tempting but ethically and legally fraught. Human oversight, clear rules of engagement and robust testing regimes should accompany any deployment of autonomous defeat mechanisms. The PLA’s push for automated swarm resilience suggests an adversary that will press the limits of autonomy. Western forces should match technical readiness with rigorous policy frameworks.
A final truth stands out from the PLA’s mixed public record. Demonstrations of resilient swarms are a warning but the PLA’s frank admission about engagement shortfalls is an opportunity. Adversaries rarely reveal their kinks on camera. When they do, allies and partners should treat those disclosures as an operational courtesy and accelerate the fielding of integrated, affordable layered defences that combine sensors, effectors and doctrine. The window to shape the next phase of the swarm competition is open precisely because real-world testing has exposed seams in both attack and defence chains.
The technical race will continue. The better strategic posture is simple to state and hard to execute: buy the layers, practise against real swarm profiles, harden and disperse assets, and keep humans squarely in the loop where lethal force is concerned. Doing so will not banish the swarm threat overnight but it will restore the defender’s leverage in the cost-imposition game and make any large-scale drone blitz a far less attractive gamble for Beijing or anyone else who tries it.