Generative Artificial Intelligence: A Transformational Catalyst for Modern Warfare and PLA Joint Operations

Abstract

This paper explores the People's Liberation Army's (PLA) perspective on generative artificial intelligence (AI) as a transformative force in modern warfare and joint operations. It argues that the increasing complexity of contemporary battlefields necessitates a shift from purely human-centric intelligence to leveraging generative AI for comprehensive support across all phases of conflict: pre-war, during war, and post-conflict.

Before a conflict, generative AI, coupled with "digital twin" technology, will revolutionize war rehearsals by creating highly realistic, dynamic virtual battle environments. This enables advanced simulation of adversary actions and optimization of joint combat plans. During operations, generative AI will be crucial for real-time battlefield perception, integrating into a military Internet of Things (IoT) to fuse multi-source data into actionable intelligence. It will enhance command and control, optimize troop deployment, and intelligently assign tasks. Furthermore, the paper highlights AI's role in ensuring battlefield data protection through decentralized storage, intelligent firewalls, and automated encryption. Post-conflict, generative AI's continuous learning and self-optimization capabilities will be vital for comprehensive assessments, refining strategies, and improving overall combat effectiveness.

Beyond its direct application, the paper delves into how AI is fundamentally reshaping the very nature of modern warfare. It discusses a shift in combat timing from traditional preemptive strikes to rapid post-strike counterattacks enabled by AI's accelerated response capabilities. The battlefield scope is evolving from expansive to more "contracted" and focused, driven by precision technologies and multi-domain integration. Finally, the paper posits a transformation in the use of force from a "physical state" (deterministic, material-based) to a "chemical state" (fluid, intelligent, and information-driven), characterized by mesh configurations, multi-point synchronous linkages, full-dimensional distribution, and multi-directional coaxial operations. These trends underscore a future where combat effectiveness is achieved through self-organization, dynamic resource allocation, and a deep, multi-layered resonance across the entire command and control system, fundamentally altering traditional military paradigms.

Autors Gabriele and Nicola Iuvinale

1. Introduction

The complex nature and increasing intensity of the battlefield in joint combat, characterized by the intersection of multidimensional and multimodal information, make reliance solely on human intelligence insufficient. The People's Liberation Army (PLA) of China, also supported by international military studies, believes that generative artificial intelligence (generative AI) can provide comprehensive support and assurance for joint combat operations. The technical advantages of such AI can be effectively applied throughout the entire war process—before, during, and after—offering a powerful impetus to seize the initiative on the battlefield and gain asymmetric advantages. This analysis explores the role of generative AI in joint operations and the profound transformations that AI is initiating in modern warfare.

2. Application of Generative Artificial Intelligence in Joint Combat

The PLA anticipates that generative AI will be integrated into platforms, weapons, and equipment, with a centralized infrastructure serving as the hub for a battlefield data collection system and a military Internet of Things (IoT) architecture.

2.1. Pre-War Phase: War Rehearsals and "Digital Twin"

War rehearsals are a crucial part of the preparation phase for joint operations. Through them, joint combat plans can be tested and optimized to ensure that various tasks and weapons can cooperate effectively in real combat. Commanders can make timely decisions based on test results.

Generative AI, with its strong data storage capacity (including geographic information, weather data, historical war cases, national and military conditions), can cooperate with "digital twin" technology to create virtual replicas of complex battle environments. The digital twin not only includes static elements but also presents dynamic details visually, providing an advanced simulation platform for designing and simulating actions. Generative AI can simulate the adversary's possible actions, allowing for familiarity with their behavior patterns and adaptation of response strategies. The test field thus built is a dynamic, simulated scene, based on a comprehensive reference to reality, containing variables and interactions. Its "decentralized" deployment allows for large and small-scale imitations, improving mutual understanding and overall combat effectiveness in joint operations, as well as helping commanders uncover potential hidden risks and dangers.

2.2. During War Phase: Battlefield Perception and Data Protection

Accurate, real-time perception of the field situation is a fundamental prerequisite for ensuring battlefield advantages and achieving objectives. Generative AI can be integrated into various platforms, weapons, and equipment to realize a data collection system and military IoT architecture, acquiring multi-source information in real time for constant monitoring.

Thanks to its data fusion and modal conversion capabilities, generative AI can intelligently and autonomously filter messy data, converting effective information into a comprehensive battlefield situation map. This helps commanders improve their perception and control of the situation, providing technical support for rapid, accurate, and efficient command and control. Based on information collection, generative AI can continuously optimize troop deployment and intelligently assign tasks according to dynamic changes on the battlefield, transforming information advantages into decision-making and operational advantages.

For the PLA, generative AI builds a unified information platform and establishes sharing channels for key information and resource status, improving battlefield transparency and efficient resource allocation. This platform can rapidly process and analyze intelligence, break down information silos, and seamlessly connect frontline reconnaissance, command centers, combat troops, and logistics for a winning synergy. Furthermore, it realizes a close link between information supply and demand sides, providing a scientific basis for formulating and adjusting joint combat plans.

Battlefield data protection represents a "red line" throughout the entire combat cycle. Generative AI, in terms of data storage, can effectively integrate different data and assign weights based on value and importance. Through "decentralized" archiving methods, data can be stored in multiple nodes, reducing single-point failure risks and improving security. Regarding data transmission, generative AI can be combined with privacy measures like firewall settings, automatically analyzing database access patterns and behaviors, setting up an intelligent firewall to identify and block anomalous access and potential security threats. It can also automatically perform data encryption, independently manage dynamic keys, and accurately and efficiently verify data integrity. For data access, generative AI can dynamically adapt staff access levels, achieving "granular" access control and ensuring information security and timeliness.

2.3. Post-Conflict Phase: Assessment Assistance

Post-conflict assessment is an indispensable part of the entire joint operations chain. Generative AI's capabilities for continuous learning and self-optimization play a significant role in improving combat effectiveness and adapting to the ever-changing battlefield environment.

Generative AI's operating mechanism includes a positive feedback learning mechanism, which can be mapped across the entire field, constantly correcting errors and improving joint efficiency. First, its algorithmic model can be continuously iterated and optimized for intelligent and refined data processing. Second, generative AI content generation, based on the joint probability distribution of existing data, can perform deductive creation after actual induction, providing useful references for solving real-world problems. Third, post-war evaluation by generative AI helps gradually shift the decision-making process from the macro to the micro-specific design, making the connection more powerful and helping decision-makers process a large number of tasks in parallel, reflecting the contingency effect in decision-making optimization.

3. New Trends in the Evolution of Modern Warfare

The continuous application of cutting-edge technologies, particularly AI, in the military field is fundamentally changing key aspects of warfare, such as combat timing, battlefield scope, and the use of force. This requires a deep understanding of these new trends to build superior combat capabilities and maintain the initiative.

3.1. Combat Timing: From "Before" to "After"

In traditional warfare, a surprise attack was a key method for gaining relative advantage. However, in modern warfare, the development of reconnaissance and perception technology and the emergence of intelligent entities on the battlefield (Military IoT) exponentially improve battlefield response times, jointly promoting the transformation of combat opportunities from preemptive to post-strike counterattack.

Battlefield situation control will make surprise attacks easier to detect, and troops will begin combat preparations while effectively responding to the opponent's attempts, significantly reducing the effectiveness of preemptive strikes. AI-based command and control links react more quickly, rapidly accumulating energy and responding effectively after detecting an opponent's surprise attack attempt. Preemptive strike in modern warfare refers to the development and use of data based on intelligent technology, leveraging the last-mover advantage to obtain more data, relying on computing power architecture for information sharing, and relying on algorithm-based decision-making advantages to form operational advantages and gain battlefield superiority.

Preemptive attack primarily takes two forms:

1. "Counterattack in danger": Based on intelligent decisions made by "humans outside the loop" (human-on-the-loop), commanders make quick decisions and rapid actions based on plan libraries and algorithms.

2. "Act on the benefits": With a human-machine decision-making process mainly based on "humans involved" (human-in-the-loop), commanders perceive battlefield changes in real time, detect important signals, and precisely understand development trends, seizing favorable opportunities like turning points in the battlefield situation.

3. 
   

3.2. Battlefield Focus: From "Expansion" to "Contraction"

Traditional warfare emphasized expanding the battlefield into multiple dimensions to compensate for threats and support independent operations. However, with the rapid development of information, intelligence, and unmanned vehicle technologies, the balance has shifted, and battlefields have begun to shrink and develop.

Command and control, mobility, and attack capabilities are shifting towards extremely fast, extremely distant, and extremely precise targets. The advancement of military technology has caused a "shrinking effect" in multi-domain and stratified space: multidimensional space has been integrated and compressed, and "thousands of miles away" have become "reachable." Troops distributed across different combat spaces have the conditions for efficient integration, developing capabilities such as holographic perception, instant decision-making, and immediate attacks.

The modern warfare battlefield will be a multi-domain integrated space, formed by the interconnection of widely distributed "points" and three-dimensional "line" electromagnetic fields. Strategy, campaigns, and tactics are compressed into a single spatial element. This will break territorial boundaries, with unmanned combat platforms capable of maneuvering beyond terrain limitations. Modern warfare presents itself as a situation of a "large battlefield with a small objective." The focus will be on point spaces such as important objectives, key nodes, and vital parts. Precision control will become the main style.

3.3. Use of Force: From "Physical State" to "Chemical State"

The traditional use of force emphasizes the "physical state," based on external properties of force and stability. Modern warfare, with the intelligent upgrading of military equipment and fast large-area networking combat modules, promotes the way force is applied from the "physical state" to the "chemical state."

The "chemical state" force application model refers to a combat module uniformly distributed over a large area, with a flexible network as an organizational structure, cloud symbiosis as information flow, an algorithm-based decision-making model, and the generation of efficiency through nonlinear emergence. Modern warfare evolves from deterministic control to an uncertain game between complex systems. Troops and equipment equipped with intelligent modules will continuously learn, improving their perception, decision-making, and attack abilities. Intrinsic properties like data, algorithms, and computing power are gradually becoming important factors for force application. The transformation from the "physical state" to the "chemical state" is a qualitative change from the superposition of single factors to the deep integration of multiple factors, and from the dominance of material energy to the dominance of informational intelligence. It is a leap from static combat effectiveness assessment based on the Lanchester equation to the emergence of a combat effectiveness system supported by Metcalfe's law.

3.4. Mesh Configuration, Multi-Point Synchronous Linkage

In intelligent warfare, combat units and functional nodes scattered across the battlefield aim to create non-linear spatio-temporal relationships for combat operations. Based on large-scale battlefield information networks combining high transmission capacity, high speed, low latency, agility, high reliability, and no loss, they autonomously form countless fractal networks through modular groupings and plug-and-play functionalities. Aided by "multiple algorithms + high computing power + excellent computing data," they predict battlefield situations and combat trends in advance, making it easier for units and nodes to precisely understand linkage timings, flexibly adapt linkage strategies, and autonomously negotiate linkage actions.

Combat operations at various points, locations, and times transform into self-organized and self-cooperative behaviors anchored to key objectives, radically changing the traditional model of achieving combat power concentration through physical massing of firepower and manpower. Instead, it relies on real-time information perception, on-demand data distribution, high-level situation sharing, and intelligent task assignment to achieve physical dispersion of force and logical concentration of effectiveness. Thanks to organic, real-time, and multi-point coupled connections, rapid, agile, and efficient movement of the entire combat system is ensured, achieving the best system confrontation effect of mobile concentration. For time-sensitive, high-value, high-risk "window" tasks, relevant forces are immediately concentrated, and synchronized actions are undertaken spatially, inter-regionally, and with time jumps. This multi-point synchronous linkage integrates flexibly configured combat units and functional nodes in different battlefield spaces, effectively unifying combat resources and achieving instantaneous penetration and omnidirectional assaults on key directions and objectives. It can create a significant asymmetric advantage over the enemy, making it impossible for them to defend, control, or resist in time, and even producing a system collapse effect.

3.5. Full-Dimensional Distribution, Multi-Domain Articulated Linkage

In intelligent warfare, the combat space covers the physical, information, and social domains, showing distinctive characteristics of inter-domain connection, multi-domain integration, and mixed-domain warfare. Building an intelligent combat system involves a large aggregation of various resources beyond the military scope and a large concentration of widely distributed elements across the entire battlefield. Through aggregation and concentration, combat functions gradually overlap, and system combat capabilities accumulate. Confrontation results and their interactions in any field can have an uncertain impact. Therefore, relying on the overall national strength is essential to realize a robust overall organization from strategic, joint, and overall perspectives, in order to constantly generate and improve the overall game strength.

Through precise docking, accurate coordination, and precise dispatch of military and local systems, structural barriers in inter-military and local operations can be gradually eliminated, effectively filling "gaps" and "lacunae" in the joint combat system and promoting accelerated transformation from loose multi-domain cooperation to tighter coordination, forming an articulated linkage similar to a metal zipper for maximum integration and cohesion. For example, camouflage has evolved from a combat support measure to an important combat operation requiring participation from all services. Its content, objectives, mission space, technical means, engineering measures, and tactical/technical requirements are significantly different from traditional camouflage. It has become a vital part of battlefield confrontation, permeating all aspects of intelligent warfare. Commanders must strengthen overall planning and meticulous arrangement of camouflage in peacetime, promoting deep cooperation and coordinated actions between military and local systems to truly "hide under the nine earths" and "move above the nine heavens."

3.6. Operational Traction, Multi-Directional Coaxial Linkage

In intelligent warfare, the two sides are in multidimensional battlefields such as land, sea, air, space, network, and electricity. Intelligent combat platforms have broken the physical limitations and geographical separation of traditional combat platforms, allowing combat power to extend extremely wide, high, and deep, and to achieve instantaneous real-time response and action, greatly blurring the spatio-temporal boundaries of the battlefield. The traditional battlefield contact line, force gathering points, and front-rear divisions are gradually disappearing. The battlefield is rapidly developing in the two directions of "unlimited expansion of range" and "high concentration of combat space." Offensive and defensive operations may no longer have a fixed "focus," and combat power will be released extremely rapidly, with extremely frequent combat conversions.

The "tentacles" of combat forces will cover the entire battlefield. As long as a mission requirement exists and is realistic, these "tentacles" can rapidly extend to any point, tangible or intangible, making it difficult to clearly define the "area of responsibility" for each combat force. Taking fire attack operations as an example, once an "order" task is assigned, it must come from a coordinated, multi-dimensional, multi-directional, and multi-directional attack force, whether manned or unmanned. It is no longer confined to the traditional long-term "discover-guide-attack-evaluate" combat cycle but relies on the empowering support of the intelligent kill chain, fully leveraging its advantages of wide scale, high dispersion, and strong saturation, performing energy release via "coaxial" linkage. This optimizes and determines in real-time the attack direction, target, order, method, and intensity of each attack platform, as well as task allocation, combination form, and path planning among multiple "moving" platforms, to achieve the optimal attack capability of "whoever discovers, whoever strikes." Combat forces originally belonging to multiple combat spaces must deconstruct, transfer, or transform space, concentrating combat energy on a relatively small spatial area to form a new spatial relationship and combat structure, and then decouple once the task is completed. Thus, the entire combat space is always in a state of dynamic drift. If this "coaxial" action targets multiple objectives, it can not only increase attack effectiveness and compensate for the unnecessary consumption of traditional attack methods (where "gains do not outweigh losses") but also "dilute" the defensive density of the system and increase the difficulty of an "unprecedented" confrontation with the enemy, taking the art of target selection and attack to the extreme of "achieving the desired effect with minimum risk, minimum time, and minimum resource consumption."

3.7. Bidirectional Penetration, Multi-Layer Resonant Linkage

In intelligent warfare, the battlefield situation changes rapidly, and the battle process enters the "countdown" era. The temporal trend of battlefield perception, situation analysis, plan formulation, effect evaluation, and feedback adjustment is extremely compressed. The number of combat units and functional nodes is growing exponentially, resulting in extremely complex combat command procedural levels and interactive relationships.

With the aid of intelligent technology's evolution and penetration into the military field, based on an intelligent command and control network integrating "digital network + information network + brain network," commanders at all levels and command agencies can be flexibly authorized to conduct analyses, evaluations, and decision-making in different locations simultaneously. The development of intelligent technology promotes the acceleration of autonomous intelligent decision-making for individual soldiers and equipment, gradually realizing a high degree of shared cognition from top-down and bottom-up. At the same time, thanks to a multi-functional, high-capacity battlefield communication network, every combat unit and functional node can communicate vertically, horizontally, and in all directions; not only can adjacent levels be linked, but also across multiple levels, gradually enabling a true top-down and bottom-up information exchange. This will produce an enormous, resonance-like response to various changes in battlefield situations, demonstrating strong combat capabilities.

The reconstruction of this combat command mode involves relying on intelligent command and control systems, deeply opening command links, flexibly implementing parallel operations, transforming the traditional vertical serial system into a bidirectional parallel system, and transforming the original periodic business processing into a real-time online intelligent system. This largely eliminates non-value-added services in the traditional command process, and the various types of business activities and their combined connections in the new combat command process are more reasonable and fluid. Based on this, commanders and command agencies at all levels can take the initiative to maintain an overall view of the situation, comprehensively consider problems, and think about countermeasures as a whole. For example, engineering combat support, as a fundamental part of the joint combat support system, can explore establishing an engineering support command mode that combines a top-down, multi-dimensional, gradual, and bypassing approach, ensuring that, in any mission situation, it can quickly establish a command relationship with affiliated support units from a vertical perspective, focus on the support needs of important objectives, directions, positions, and seasons, and ensure timely and efficient engineering support to various combat forces. At the same time, engineering support command organizations at all levels can also make independent judgments and make immediate adjustments based on the combat situation and support effects, keeping pace with support objectives and collaborative units. They no longer rely exclusively on top-down command control but instead achieve accurate and efficient engineering support through self-organization, self-adaptation, and self-coordination, ensuring the continuous combat capability of key support objectives.

4. Conclusion

The PLA's vision for the integration of generative AI into joint combat operations and the trends in modern warfare illustrates a paradigm shift. AI is no longer merely a supporting tool but a catalyst reshaping the fundamentals of military strategy and tactics. From predictive war simulations and real-time data protection to the redefinition of combat timing and force concentration, AI is projected to create a more agile, responsive, and interconnected combat environment. The transition from a "physical state" to a "chemical state" of force application, the formation of self-organizing "mesh" networks, and multi-layered resonant command underscore a future where operational effectiveness hinges on the synergy between human intelligence and autonomous AI capabilities. This evolution propels armed forces to develop integrated, intelligent combat systems to maintain a strategic edge in the rapidly transforming battlefield.

Download paper

Generative Artificial Intelligence_ A Transformational Catalyst for Modern Warfare and PLA.

Scarica • 471KB