Will it be seen in the next war that machines do these things without supervision and direction on the part of humans?
A recent article from The New York Times with sage advice and warning, “Overmatched,” highlights that the progress that is being made relating to autonomous swarms of drones is getting beyond what the Pentagon can handle. Simulations done on war gaming against China have shown U.S. forces as wanting and vulnerable to what an enemy would be quick to exploit.
It’s no distant vision that Ukie clever tricks on today’s front, China’s swarming tests, and AI Strike Engines being simultaneously developed by Western companies will soon merge into a battlespace wherein size, autonomy, and agility will be more prized assets than weaponry. The stakes here are stratospheric. A loss of defining advantage for the USA would appear imminent without an air of crises and planning.
1. The Pentagon War Games: Continuing Losses
“Overmatch brief” can be seen as an internal report given out by the Pentagon, as per NYT, and it shows they have been on the losing side in all the simulations they conducted against China. The reason for it is structure. Their weapons are assembled with multi-million-dollar technologies like F-35s, and journalists get assigned dirt-cheap and flexible tools. It’s very easy to see why Bob Work, who once served as the Deputy Defense Secretary, spoke about Chinese commanders’ abilities to disrupt American battle networks via jamming and cyber attacks back in 2021.
2. China’s Advances toward Fully Autonomous Drones
The Air Force within the People’s Liberation Army in China has shown swarming capabilities with autonomous actions. It is not yet involved in combat operations, but it is running very low on time before it will be. It gets the benefit of military AI, DeepSeek, which analyzes thousands of scenarios on a battlefield within seconds. It would be observable here that these swarming networks, once deployed, might be capable of attacking enemy defenses without any intervention on an operational level.
3. Mass Semi-autonomous Drone Warfare – Ukraine
“Today, 9,000 drones are launched daily. Local production of FPV drones is 200,000 per month with an annual production capacity of 4 million drones. Use of AI modules reduces costs to 70 US dollars. Its accuracy rate enhances from 20 percent to 80 percent. Fiber optic tethered drones, which are not affected by interference and comprise 10 percent of drones on hand today.”
4. Battlefield Innovation under Adversarial Conditions
The drones have been employed in battles that have been devised based on environmental and operational considerations. When there are heavy fog and rain conditions, ground drones help reactive drones on intercept missions against cars operating based on environmental deception. Deception on operations employs drones fitted with speakers that lure Russian forces into targeted locations. Unconventional methods, which include exploding a failing drone and causing a structure to collapse, comprise the ingenuity that leads to Ukrainian success. Perhaps an additional critical advantage within drone attacks might be adaptability or readiness.
5. Western AI Swarm Systems on Front Lines
Nemyx engine from Auterion, launched in September 2025 under an alliance with the Swiss-American company Auterion, links drones provided by different suppliers as per coordinated attacks controlled and led by artificial intelligence. Skynode AI kits numbering more than 33,000 are on their way to Ukraine with agreements with the Pentagon, and it appears that all these components are upgradeable with Nemyx capabilities. Anduril’s Fury drone achieved its first successful mission using artificial intelligence guidance in October 2025 as per an overall plan for 1,000 AI wing drones. These developments have clearly shown that swarming systems can be assembled effectively within Western industry capabilities, but it’s still unclear at which rate they can proceed compared with developments within the opposing side.
6. The Counter-Drone Challenge: No Silver Bullet
It has also been commented that swarms would be effectively defeated only with a multi-layer defense. Directed energy weapons such as high-power microwaves would have the capability to destroy dozens of drones at once, with Epirus’ Leonidas system said to have destroyed 49 drones within less than a second, but these would be limited. Already, electronic warfare would have multi-frequency links and optic fibers. It would be impractical to have a large number of kinetic interceptors on account of inexpensive drones. The projects launched against the Pentagon would produce an offspring of offense and defense drones on an industrial scale but would be very delayed compared with China.
7. Command and Control Enabled by AI
Having AI interoperability for battle management systems will be essential for winning against swarms. Machine learning will be useful for helping to accelerate target detection and rapid response. Examples include Palantir’s Maven Smart System, which integrates domain-to-domain data. All these will enable battlespace understanding and thus rapid sensor-to-shooter operations. A lack of AI interoperability for C2 would mean that manned processes within C2 will be bottlenecks and would be unable to sustain autonomous operations rhythm.
8. Size and Responsiveness of Industry
It would appear that the drone warfare arena sets forth on an endless competition with regards to innovation and response. It would appear that Ukraine’s designs and very short cycles performed on a 4-6-week level differentiate considerably compared with a relatively slow procurement process that took place within the Western sector. It would appear that it sets forth on an arena requiring the strength of productivity on a large scale and readiness within designs toward immediate responses on a real-time scale. There would be no chance for forces with less adaptive productivity cycles and software.
9. Risks of a Call for Arms Control as Escalation Escalates
The NYT editorial corresponds with UN and ICRC calls for a 2026 Convention on autonomous weapons. Andrew C. Weber writes, “The pace of warfare will soon be too fast for humans to control.” The irony here for the U.S. is that it develops and proliferates superior autonomous weapons systems and also begins to advocate a ban on consumer drones Dji Mavics that have no applicability to military swarming. Without an integrated strategy, arms control agreements could be left mired and jeopardized by progress. The crossroads and consolidation of mass production, autonomous intelligence via AI, and adaptational methods will continue to cloud and cloud warfare on several fronts.
Rapid developments on Chinese swarms, Eastern Europe’s breakthroughs on Ukraine’s battlefields, and current developments today within Western capitals make it increasingly clear that an imminent stance on temporal cycles on warfare will soon be reached wherein victory will be measured as much on adaptational rates as on superiority. The U.S. clearly faces two dilemmas here. It must develop and integrate autonomous systems at a rapid enough rate as to have some chance at competing. It must be at the forefront on necessary global developments on arms control standards so as not be swamped on the wars ahead.
