China’s AI Breakthrough: Missiles That Learn and Adapt

China's AI-powered missiles can adapt their flight paths in real-time, posing a significant threat to global security.

Must Read

Girish Linganna
Girish Linganna
Girish Linganna is a Defence & Aerospace analyst and is the Director of ADD Engineering Components (India) Pvt Ltd, a subsidiary of ADD Engineering GmbH, Germany with manufacturing units in Russia. He is Consulting Editor Industry and Defense at Frontier India.

How might a missile “in flight” adjust to threats in enemy defense systems instantly, using AI-assisted sensing and data analysis? More specifically, what if, in a matter of milliseconds or seconds, AI-powered multi-domain sensors detected fresh information and quickly altered the path or trajectory “in flight”?  

Maybe there’s a “gap” in the enemy’s radar aperture, or a defensive perimeter weakens in a certain location, or high-value targets that were previously hidden from view become apparent. A long-range precision attack missile might be redirected by sophisticated AI-supported guidance systems that organize and process multi-domain sensor data.

These scenarios form the basis of what a new significant Army intelligence report describes as emerging Chinese concepts of multi-domain operations concerning long-range precision strikes. The Army’s Training and Doctrine Command’s G2 (TRADOC G2) intelligence service outlines this Chinese tactic or ambition in detail, referring to China’s “intelligentized C2 and ISR structure.”

The PLA is looking into more advanced long-range precision strike systems that will have a strong, networked, and in the future smart C2 and ISR structure, which will make their anti-access/area denial (A2/AD) abilities better. These are the tenets of the PLA’s main operational concept, known as Multi-Domain Precision Warfare (MDPW), says TRADOC G2 Warrior, citing its published report, The Operational Environment 2024-2034 Large Scale Combat Operations.

The G2 report, which includes a published analysis of current and anticipated future wars, thoroughly examines technological trends and innovations to best characterize the expected “operational” military environment over the next 10 years. The report explains this Chinese strategy as “Multi-Domain Precision Warfare” (MDPW).

Chinese Concept of Multi-Domain Precision Warfare

According to TRADOC G2, MDPW would use a C4ISR network that uses big data and AI to quickly find key holes in the U.S. operational system and then bring together forces from different areas to launch precise attacks against these holes.

This appears to be significant in several aspects, including long-range attacks, operationalized AI, and multidomain networks. Perhaps most notably, China’s “intelligentized” MDPW seems to be a form of emulation, aiming to replicate the Pentagon’s now-realized efforts for Joint All-Domain Command and Control (JADC2), but with perhaps a more aggressive concept of using AI for offensive, lethal attacks using long-range munitions.

The primary difference between China’s MDPW and the U.S. Department of Defense’s JADC2 seems to concern ethics, target precision, and concepts of operation.

The Pentagon and military services already use AI to transmit critical information across domains for targeting; however, executing a lethal attack requires human command and control. For instance, the U.S. Army’s Project Convergence experiment demonstrated that AI can reduce the sensor-to-shooter attack windows from 20 minutes to a few seconds.

A distinctive feature of the U.S. approach is that it prioritizes the human-machine interface, among other things. As early as 2020, FireStorm, an AI-based system, demonstrated its ability to receive incoming sensor data from disparate sources or information pools, organize it, and analyze it in relation to the overall battle picture or scenario to “recommend” the optimal “shooter” or “effector” for a given target situation.

The AI system does this by parsing incoming data from an extensive problem-solving database, making inferences, and performing analytics almost in real time. The U.S. operational concept includes “recommending” to a human in command and control, ensuring the optimal combination of machine-human interface, and guaranteeing that humans decide regarding lethal force.

It doesn’t seem that the PLA’s MDPW operates under similar parameters. The PLA’s rapid efforts to copy or recreate the Pentagon’s emerging multi-domain network heavily rely on AI for delivering lethal offensive strikes, seemingly without requiring human intervention.

Many new U.S. weapons technologies imply that munitions can adjust their course in flight or respond to new information as it becomes available. For example, the Air Force’s Golden Horde collaborative bomb program demonstrates the ability of AI-enabled weapons to gather, process, and share data “in flight” to adapt to changing target information. It may not be clear if this is in line with the PLA’s capabilities or an attempt to copy them, but it is clear that MDPW wants to use AI and high-speed multi-domain ISR to change the path of long-range precision weapons or make them more accurate.

Networked Air Force and Army Weaponry

DoD (JADC2) has already demonstrated the ability to link and transmit analyzed data across domains using common technical standards, advanced interfaces, and so-called “gateway” technologies capable of essentially combining and broadcasting data from incompatible transport layers. Perhaps certain time-sensitive threat information arrives via GPS from one domain and requires combining, organizing, and analyzing with other information coming from another domain via radio frequency or data link.

The G2 report describes the integration of ISR data and precise targeting into a long-range precision strike if these processes occur quickly, efficiently, and accurately.

However, the PLA’s approach may not reflect the technological characteristics or complexity of U.S. JADC2, and it appears to prioritize the “science” of AI over the need for “human” input. The G2 report also underscores the PLA’s preference for the “science” of AI over human input.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Latest

More Articles Like This