It is 2028. An infantry company, acting as an advanced guard, maneuvers to support an attack in the Indo-Pacific. Each platoon consists of manned-unmanned squad teams using autonomous ground vehicles equipped with sensors, mortars, and loitering attack drones. The company commander, analyzing real-time data from small reconnaissance drones, satellites, and predictive terrain models, quickly identifies key terrain to secure the flank. He assigns a platoon to the position and directs others to adjust fires in support of the main effort.
As orders flow, algorithms assess terrain, readiness, doctrine, and historical cases to generate engagement area options. These recommendations include named areas of interest and high-value target assessments based on possible enemy orders of battle. Platoon leaders refine these insights and issue rapid guidance using troop-leading procedures. Mission command is now a fusion of human and machine judgment, enabling data-centric warfare.
This vignette reflects key aspects of the Army Campaign Plan and the drive to field a data-centric force capable of what is known as “multidomain operations” at echelon before 2030. The land domain is central to generating effects across other domains, requiring leaders to achieve situational awareness faster than adversaries. Success depends on leveraging data for targeting, maneuvering, and disrupting enemy cohesion. Decision advantage — visualizing battlespace geometry and acting faster than the opponent — becomes the new theory of victory.
At the heart of this transformation is a redefined approach to mission command in the age of AI-driven warfare. Militaries worldwide are racing to integrate AI into command systems, seeking to accelerate tempo and coordinate unmanned swarms. Yet, what does this mean for mission command?
Modernization alone will not suffice. Without a deliberate shift in leader development, the Army risks falling short of its campaign plan and National Defense Strategy objectives. This approach should start with changes to leader development and how the Army educates the force. Specifically, the Army will need to ensure combat leaders understand the basics of AI and machine learning, as well as how new systems aggregate data to make recommendations. Second, the institution will need to build on ideas in the Army Learning Concept to help leaders at echelon envision the future fight while capturing data about decision-making. This feedback loop will help the Army refine algorithms sorting the stream of data and adapt to context in a manner that makes mission command possible in human-machine teams.
Mission Command: Origins and Evolution
Mission command in U.S. military doctrine enables decentralized execution, allowing subordinates to adapt to battlefield friction. The U.S. Army defines it as an approach to command and control that empowers subordinate decision-making based on competence, mutual trust, shared understanding, commander’s intent, mission-type orders, disciplined initiative, and risk acceptance. Mission command is both a collection of individual attributes and a system for adapting to change on the battlefield based on philosophy, education, and culture in the profession of arms.
The concept has evolved across generations, with each era drawing from history and technological change to refine its principles. While often attributed to the Prussian military, its origins and reference points are more diffuse in American history. For Civil War leaders and reformers Phil Sheridan and Emroy Upton, it was a merger of Prussian technology and mobilization techniques with American tactics. For many it was the mastery of Helmuth von Moltke the Elder. After World War II, Army generals including William Depuy and Donn Starry proclaimed seminal case to understand mission command shifted to how German forces fought outnumbered on the Eastern Front and employed a mix of tactics to deny breakthroughs. In more recent years, everything from capturing Eben-Emael in 1940 to the 2003 Thunder Run served as crucial cases for articulating what mission command is for a generation.
The resurgence of mission command post-2012 reflected lessons from Iraq and Afghanistan, where centralized control failed to keep pace with battlefield complexity. In 2012, Chairman of the Joint Chief of Staff Martin Dempsey released a white paper on mission command. The text set off a wave of doctrine updates and debates that culminated in the 2019 publication of ADP 6-0 Mission Command: Command and Control of Army Forces.
The central idea was that the basic principles of mission command, including commander’s intent, mission-type orders, and decentralized execution, required adaptable leaders. This vision is consistent with earlier treatises on command that focused on the role of individual initiative in a larger system of education. Mission command fosters initiative by relying on education and shared experiences rather than rigid procedures. It thrives in training and education environments that prioritize decision-making, risk-taking, and problem-solving over rote learning. Tactical decision games, field exercises, and mission rehearsals reinforce common reference points, enabling rapid communication and adaptation in combat.
This vision differentiates mission command from learning based on school-preferred answers and a managerial approach to leadership. This conceptualization means that mission command is as much a philosophy as a system for command and control that is based on education and shared experiences. Early critiques of the Army’s post-2012 mission command efforts highlighted bureaucratic garrison training as a constraint. Effective mission command requires not just philosophical grounding, but practical reinforcement in unit-level training and decision-making.
Training and education remain essential. Soldiers should learn how to interpret crucial cases, including reenacting them as decision games and map exercises to ground their technical expertise and enable mission command. Outside of formal education, leaders use new networks to create common understanding through activities like mission rehearsals. Collectively these activities produce tacit knowledge that facilitates rapid communication and decision-making, but they are always context-bound, thus supporting decentralized execution.
Mission Command and AI/Machine Learning
The concept of algorithmic warfare links to the emergence of battle networks and concepts like “mosaic warfare,” which envisions war as a data-centric complex adaptive system. Just as the telegraph changed the ability of land forces to coordinate dispersed formations, modern battle networks are often global and touch multiple domains and dimensions. In Chinese military doctrine, this phenomenon is one of the driving trends changing the character of war, creating a need for “multi-domain integrated joint operations.” The sheer volume and rate of information drives a need to better integrate AI and machine learning to create a common operating picture and decision advantage. In Chinese military doctrine, this is referred to as “intelligentized warfare.” All modern militaries are pursuing the ability to build, fight, and protect kill webs — often cross-domain — at machine speed faster than their adversaries. This race depends on data, infrastructure, and talent to field new technologies and the primary algorithms that drive decision advantage.
These new technologies and the resulting mosaic kill webs don’t replace mission command. Rather, they augment it. Soldiers find new ways of articulating intent and delegating execution through algorithms to human-machine teams. Taken to its logical conclusion, this implies a soldier could express the commander’s intent to a machine (i.e., autonomous swarm) as long as the algorithm had a common set of contextual reference points with the soldier. The soldier would state the objective and expanded purpose, including any limitations and preferences for the algorithm to execute. This process could include feedback, with statistical analysis providing assessments of the alignment of ends with means and the probability that the mission can be executed given current readiness.
This interaction implies that the Army should have a living inventory of contextual references as data that can facilitate the communication of intent. Just as map exercises, staff rides, and wargames historically served as forums for honing creativity, judgment, and operational art, there is a need for new forums that produce tacit data for human-machine teams.
As technology changes, the education system should adapt to create a culture of learning that helps soldiers take advantage of new systems and develop new tactics and missions. This adaption should capture and code data to enable all parties — human and machine — to build an inventory of conceptual reference points.
Adapting the U.S. Army to an Era of Algorithms
First, the Army should codify how it intends to capture, process, store, and disseminate data from the tactical edge, across the theater, and back to sanctuary. Gaining and maintaining decision dominance is predicated on the ability to leverage large, diverse, and authoritative datasets. The Army should be clear on what types of data are important for its warfighting and business systems, particularly in a resource-constrained environment. With policies governing data provisioning, replication, and disposal across battlefield and theater nodes, the Army can ensure the right data is available at the point of need, mirroring its logistics enterprise. Soldiers would not only make algorithm-driven decisions at critical moments but also enable dynamic reprogramming — both at the tactical edge and in sanctuary — as adversaries adapt tactics and waveforms.
Second, the Army should reinvest in professional self-development and education along the lines mapped out in its new learning concept. Too often, existing online training does not directly address tactics and basic professional competencies. Soldiers should be required to learn common tactics and operational understanding, regardless of their specialty or whether they are in the active force, guard, or reserve component. Tactics and understanding how the Army fights should be a foundation and yet, there is no required online training or learning curricula that allows soldiers to demonstrate competence much less pursue mastery. Lethality starts with knowing how the Army fights. Worst still, there are few opportunities for tactical decision games incorporating the use of AI. The learning concept offers a vision for addressing these deficiencies but the Army — and Congress — will need to fund it to turn it into a reality and a foundation for data-centric warfare. This includes developing a data infrastructure that learns from its students and better tailors content to their needs.
Third, the Army should determine how it intends to train soldiers of all specialties on data literacy and algorithmic warfare. With today’s budget constraints in mind, initial steps could prudently focus on basic data literacy and familiarization with the principles of data-centric warfare — essentially a rubric of “all getting some, with some experts getting all.” While specialization within warfighting functions or branches helped the Army maintain its combat overmatch through expertise, the societal disruption caused by AI requires overlaying technical education across the profession of arms. Every soldier should understand data, how it is used in AI, and how it can be manipulated. Additional professional military education should address more complex topics and application skills for more senior leadership and staff roles. Again, the Army Learning Concept offers a vision to this end, but it will require resourcing and fundamental changes in training and education to become a reality.
Finally, as the Army changes its unit of action focus from the brigade to the division, so too should it change the nature of its training for multi-domain operations. While the profession of arms requires tough and realistic training that simulates the hardships and depravations experienced in combat, the complexity of multidomain operations will require stressing the cognitive capabilities of commanders and staff continually in garrison. Synthetic environments can be tailored to individual and collective training tasks — a capability already being used in Ukraine to hone air defense. The Army can pull data from these virtual battlespaces to build digital twins that leverage AI and available data sets to simulate the volume, speed, and consequences of algorithmic decision recommendations. The result will be constant, iterated sets and reps that help combat leaders gain familiarity with systems to trust agentic outputs and build the resiliency to withstand the tempo of data-centric warfare.
Conclusion
Realizing the goals of the Army Campaign Plan to build a data-centric force capable of conducting multidomain operations at echelon before 2030 requires developing a concept for conducting mission command through algorithms. The integration of AI and machine learning applications across the force will not replace the core tenets of mission command nor diminish the need for shared understanding, trust, initiative, and mission-type orders (i.e., centralized planning, decentralized execution). Rather, building a data-centric Army requires visualizing and describing how to execute mission command as a dialogue between man and machine. This process starts, like the history of mission command, with education.
Benjamin Jensen, PhD is the Frank E. Petersen chair for emerging technology at the Marine Corps University School of Advanced Warfighting and a senior fellow in the Futures Lab at the Center for Strategic and International Studies. He is also a reserve officer in the U.S. Army.
Maj. Gen. Jake S. Kwon is the director of strategic operations in the Headquarters, Department of the Army G-3/5/7. The views expressed are those of the authors and do not reflect official positions of the U.S. Army and/or Department of Defense.
Image: Midjourney
Jesus Christ is King
