Editor’s note: Last year, Army Futures Command’s Directorate of Concepts announced an essay contest to generate new ideas and expand the community of interest for the Army’s next operating concept focused on 2040. Contest entries were invited to respond to the following question: With AI maturing, autonomous systems and robotics becoming more prevalent on the battlefield, and battlefield transparency increasing, how should Army forces operate, equip, organize, and array the battlefield 2040 to overcome those challenges? This entry, from Captain Matthew Moellering, Captain André Michell, and Captain Taylor Michell,, was selected as the runner-up.
The accelerating maturation of artificial intelligence will shatter current notions of the character of war. Recent strategic guidance from the National Security Council, Department of Defense, and Secretary of the Army highlights the Army’s organizational commitment to leading global AI development. By the year 2040, next-generation AI-enabled systems will radically obfuscate the battlefield, disrupt traditional ideas of power, and fundamentally alter war. AI is the primordial technology guiding this change during the information age. To win in 2040, the Army must organize its forces and develop the necessary infrastructure to enhance cooperation with a civilian sector that has outpaced it in AI development.
As Chairman of the Joint Chiefs of Staff General Mark Milley said at the West Point class of 2022 graduation, AI has the potential to revolutionize the character of war. Most radically this technological revolution has the potential to alter where and how wars are fought. The battlefield in 2040 will encroach upon the homeland and extend beyond the forward line of troops as information warfare becomes tightly integrated with combat operations. Increasingly effective methods of weaponizing information provide a new avenue for adversaries to practice Sun Tzu’s “supreme art of war”—“to subdue an enemy without fighting.” Beyond a dynamic battlefield, AI offers an asymmetric capability for underestimated, or underfunded adversaries that can disrupt traditional notions of air or naval supremacy. An unmanned underwater system or unmanned aerial system will provide localized air or undersea superiority that negates or extends the previous sources of military power from highly advanced aircraft or ships. The Army of 2040 will contend with a broader, more obfuscated battlefield and more dynamic enemy capabilities than ever seen before. The current data landscape of ad hoc organizations provides capabilities at inappropriate echelons with poorly aligned authorities and responsibilities. The Army will only ensure its operational ability to fight in this data-rich environment by synchronizing its data workforce with the authorities and structures commensurate to its role in future military operations. A Data Warfare Regiment, consisting of both operational data science teams and an Army Data Corps is necessary to synchronize federal resources and develop the national AI infrastructure to win in 2040.
Modern conflicts such as the 2020 Nagorno-Karabakh War and the ongoing war triggered by Russia’s invasion of Ukraine have demonstrated the devastating effectiveness of AI on today’s battlefields and proven the willingness of other countries to use these tools in conflict. Unmanned, AI-enabled weapons such as the TB2 Bayraktar have been effective against both Armenian and Russian armored formations, making combined arms maneuver more expensive and difficult. At the same time, operational planning tools like Project Maven have enabled American planners to more effectively target and eliminate enemy forces. These weapons and systems point to a growing tidal wave of technological innovation primed to disrupt how we wage war.
Skeptics point to unreliable effects from military applications of these technologies across diverse battlefields to question the importance of committing to restructuring the Army to lead AI development. Their arguments are buoyed by inappropriate and inefficient applications of a nascent technology. Consider how the use of unmanned systems might mirror the adoption of the machine gun during the industrial revolution. Fifty years before the machine gun contributed to the most deadly battles seen up to that point in human history, the weapon system was widely fielded in the Franco-Prussian War. It was initially applied as a form of artillery, arrayed within artillery batteries, and had limited effects during the war. It took revolutionary reorganization of infantry forces and reconsideration of modern maneuver to fully take advantage of automatic weapons’ devastating effects. AI-enabled systems are prone to similar misuse today by an Army too often criticized for preparing to fight yesterday’s war tomorrow.
As the US Army of 2040 plans to be more autonomous, and China prepares to become the world’s preeminent AI superpower by 2027, the Army must consider successful applications of autonomous and AI-enabled systems like the Bayraktar and Project Maven for effective competition with China. Although some units have utilized AI, the current organizational structure is insufficient to fulfill the unique requirements of an AI-enabled Army of 2040, optimized for Multi-Domain Operations. AI demands massive amounts of data and a specially educated workforce for its effective implementation. As disruptive as this technology might be, preparing for it has proven difficult but highly effective. By identifying opportunities in existing data-rich environments, soldiers have been laying the groundwork for a reorganized Army arrayed to take advantage of immediate benefits for AI-enabled systems and weapons at the intersection of intelligence and operations.
Operational Data Science Teams on the Battlefield
There are several examples of success in force structure modification in recent history. During the post-9/11 wars, JSOC synchronized intelligence and operations planners to enable faster, more effective operations. As the amount of data collected and the complexity of analysis both increase, traditional intelligence functions have needed support from statisticians and complex systems to accomplish commanders’ objectives in combat. Largely, this support has been unavailable and large amounts of data that can drive intelligence is frequently underutilized.
Operations research systems analysts have a long history of providing support for the Army’s statistical modeling and economics analysis needs. Some trained in machine learning have even served on operational data science teams during various missions, within special operations units, and on staff at echelons at and above brigade. Due to their diverse skill sets and backgrounds though, operational research systems analysts with the skills to operationalize intelligence with modern AI techniques are rare and lack the organizational support for a career in the Army. There have been exceptional success stories and individual dynamic leaders have had an outsized impact, but these successes are not sufficiently scalable or ubiquitous to offer success on the battlefield of 2040. The Army requires new structures for leaders to be successful on the AI-enabled battlefield of 2040.
The Engineer Branch: A Model
When considering structures for an AI work force many have looked to the Army’s cyber, signal, and intelligence branches. The Engineer Regiment provides a better framework for the scale and scope of the capabilities gap the Army must fill, making a strong case for the establishment of a Data Warfare Regiment. A key characteristic in both the Engineer Regiment and the proposed Data Warfare Regiment is the distinction between operational units and the corps that focuses on public works. Imitating the Engineer Regiment would enable coherent integration with operational units, collaboration with industry, and rapid domestic modernization, all of which would better equip the United States to compete with foreign powers. The Army Corps of Engineers’ mission—to “deliver vital engineering solutions, in collaboration with our partners, to secure our Nation, energize the economy, and reduce disaster risk”—resonates with the Army’s needs for AI development. The Data Warfare Regiment can use the Engineer Regiment as a guide in education, disaster response, and relationship building with industry.
The Army Corps of Engineers was founded in 1802 in conjunction with the United States Military Academy and brought “science into government.” Education became tightly intertwined with the Corps of Engineers’ mission as the superintendent of the US Military Academy was an engineer officer up until 1866. Engineer officers are not only required to get a master’s degree for career progression, but are also encouraged to receive education in nine engineer-specific additional skill identifiers and to pursue professional engineering licenses. This closely resembles the Army’s initial approach to AI talent development and management. While the value placed on graduate education has spread to the rest of the branches within the Army, the Army itself has yet to institutionalize graduate education in technical fields. The Air Force created the Air Force Institute of Technology to ensure “technical and professional continuing education.” The Navy created the Naval Postgraduate School to deliver “emerging capabilities at speed and scale.” Because the Army does not have a comparable education infrastructure, setting up Army-led higher education focused on AI and related technologies would be the first charge of the Army Data Corps. The Army Data Corps, in a similar spirit to the Corps of Engineers, would then look to build tools to improve AI robustness and fairness, and to secure the home front against AI-enabled attacks from our adversaries.
The Data Warfare Regiment
Winning on the obfuscated, dynamic battlefield of 2040 will require a multifaceted approach to organizing technologically advanced units. It may seem paradoxical to describe the AI-enabled battlefield of 2040 as obfuscated as the Army prepares to be more data-oriented, equipped with advanced analytics and operating with increased situational understanding. As we prepare to fight large-scale combat operations, however, our data-reliant, AI-enabled systems will be challenged to keep up with the limitations imposed by denied, degraded, intermittent, and limited combat networks. When data is therefore sparse or unavailable, it will either incapacitate a system or leave the unavailable data to be imputed by advanced analytics. These analytics come to uninterpretable predictions, communicated in probabilities that people struggle to understand. Furthermore, building AI-enabled combat systems greatly increases their complexity and breadth. An AI-enabled combat system will be infinitely more complex than its modern counterparts, will be vulnerable to confidentiality and integrity attacks from near-peer adversaries’ cyber forces, and will spread the battlefield deep into the strategic support area.
The Data Warfare Regiment would serve in the conflict zone—the area of ongoing kinetic operations—through data warfare teams and on the homeland through the Army Data Corps. This organization mirrors the previously described Engineer Regiment in its integration with operational units and collaboration with technological pioneers at home. By maintaining a presence with industry, the Army positions itself to leverage our national advantage in AI. Positions within operational units provide operational relevance for future applications of these technologies. This structure further provides the flexibility for service members to build successful, fulfilling careers in this field.
The data warfare team enshrines the ad hoc operational data science teams created over the last two decades into structures attached to each operational unit. An AI-enabled battlefield necessitates an organization at the edge to maintain, build, and employ AI models and autonomous systems. While AI offers automation and efficiency benefits, it requires significant training and an openness to new ideas and methods. In an organization with high turnover that struggles to maintain systems beyond a single commander’s tenure, a data warfare team will be able to customize and tailor solutions for commanders and their staffs. This tighter integration of technologists and warfighters will build trust in critical systems institutionalizing a support function specifically for operationalizing data.
On the dynamic, highly dispersed battlefield of 2040, small units will simultaneously be critical to mission success and incredibly vulnerable. While the Army is already capable of massing effects on enemy targets with devastating effectiveness, it will similarly need to mass information to drive operations. Modern data science and AI-enabled systems are reliant on large, networked services provided by cloud service providers. This has enabled private companies to adopt work-from-everywhere policies as the computing resources are similarly dispersed. As the battlefield expands, the need for localized support to operational units will become even more important. Should the battlefield of 2040 have even a fraction of the sensors predicted, the amount of data generated will be too great to reliably transmit over combat networks. To effectively adapt AI for the environments the Army will fight in, technologists will need to be integrated directly with the units they support. The data warfare team will have the personnel and capabilities to operate mobile data centers that support local AI and autonomous systems.
The Army Data Corps protects the home front by contributing to national security priorities and providing a local home for Army technologists to collaborate with industry partners. In this organization, service members can enrich their military careers with educational assignments within academia and our nation’s critical industries. In 2040, this organization will be tightly coupled with the highly successful and influential open-source software and AI development community. By contributing to open-source systems, the Army Data Corps will strengthen a sometimes strained relationship with the tech industry. The Air Force’s contribution to DevSecOps demonstrates the potential for an Army Data Corps to support software solutions that bolster the security and resilience of national infrastructure. Contributing to the tools Americans rely on for their information infrastructure supports a competitive economy and ensures American companies can continue to compete at home and abroad.
Assignments within the Army Data Corps would serve two functions for a service member’s career. Firstly, it would allow technologists to hone their skills in support of national security objectives. This combination of technical challenge and meaningful work will be highly effective in retaining the talent the Army seeks to attract and maintain through 2040. Secondly, it will enable service members to refresh skills in a rapidly evolving field by collaborating with already established organizations like the Army’s software and AI factories. These experiences will in turn prepare service members to support data warfare teams with fresh ideas and new methods to help the Army continue to fight and win our nation’s wars. Active involvement with the broader tech community could also inspire support from unaligned communities like Bellingcat and other hacktivist organizations in future conflicts.
Reorganizing for Tomorrow’s War
The changing character of war requires the Army to reorganize its forces to prioritize structures that enable data and decision dominance at the speed of relevancy. Anticipating a radically extended battlefield, the Army will need to mass information to drive operations. This will bring AI and software engineers onto the battlefield in a new way. The Data Warfare Regiment will provide a career path for these individuals and support the nation’s growing reliance on information infrastructure. Collaboration with the tech industry at home will provide a democratic alternative to our adversaries’ autocratic employment of AI-enabled systems while supporting the decision dominance of our Army.
Captain Matthew Moellering, US Army, currently serves in the second cohort of Army Artificial Intelligence Scholars at Carnegie Mellon University. He is pursuing a master of information systems management / business intelligence and data analytics from Heinz College. He is a Special Forces officer with deployments to the Middle East and Afghanistan. He graduated from the United States Military Academy with a bachelor of science in mathematics in 2014.
Captain André Michell, US Army, serves as a data engineer at the Army Artificial Intelligence Integration Center in Pittsburgh, Pennsylvania after completing the inaugural class of the Army Artificial Intelligence Scholars program at Carnegie Mellon University. Michell is an infantry officer who has served as a mortar and heavy weapons platoon leader in 1st Battalion, 27th Infantry Regiment. He holds a BS in computer science from the United States Military Academy and an MS in computational data science from Carnegie Mellon University.
Captain Taylor Michell, US Army, serves as a data scientist at the Army Artificial Intelligence Integration Center in Pittsburgh, Pennsylvania after completing the inaugural class of the Army Artificial Intelligence Scholars program at Carnegie Mellon University. Michell is an engineer officer who has served as a sapper platoon leader in the 65th Brigade Engineer Battalion. She holds a BS in chemical engineering with a minor in grand strategy from the United States Military Academy and an MS in information systems management with a focus in business intelligence and data analytics from Carnegie Mellon University.
The views expressed are those of the authors and do not reflect the official position of the United States Military Academy, Department of the Army, or Department of Defense.
Image credit: Master Sgt. David Eichaker, US Air National Guard