15. The Real AI Advantage: Integrating Leadership Decisions Across Military Information Domains

Framing the Challenge

(The gradual, continuous adaptation of military organizations, technologies, and doctrines over time in response to changing strategic, political, and technological conditions.)

Senior decision-making is central to all military operations. Using AI to inform, accelerate, and support these decisions continues the longstanding practice of leveraging technology for decision advantage. Today’s AI-enabled decision support is evolutionary: it is not fundamentally new, but it is faster, more accessible, and more detailed.

In a meeting I attended, a team discussed the data and assumptions needed for an AI-enabled simulation to support a complex, multi-step decision process. The goal was to produce repeatable outcomes much faster than existing methods. The simulation produced detailed results in 30 to 45 minutes, whereas legacy systems required weeks to produce similar outputs. This case illustrates the substantial efficiency gains achievable through AI, highlighting how technological integration can drastically accelerate complex decision-making cycles that were previously constrained by slower, non-AI automated processes.

Notably, a subject matter expert (SME) using only a pen, a napkin, and years of experience, processed the same inputs and reached a comparable answer in about 15 minutes, within ten percent of the AI’s result.
This comparison underscores a key aspect of military decision-making. Senior leaders trust experienced SMEs because their judgment has been proven over time. However, such expertise is rare and takes decades to develop. It cannot be easily replicated or scaled across all decision domains. Individuals with broad, reliable judgment are even less common.

AI-enabled decision systems address this challenge by making expert-level analysis available across the organization, rather than limiting it to a few SMEs. AI-generated results are consistent, repeatable, and rapidly produced. They are accessible to a wide audience and can be shared, rerun, and adapted as needed. While AI does not replace the intuition of a seasoned SME, it captures and disseminates aspects of expert decision-making logic at scale that human expertise alone cannot. The SME’s napkin was useful for illustrating reasoning to a small group, but it could not support communication or execution across a large organization. Well-designed AI simulations can fulfill that role.

It is important to recognize the limitations of current AI systems. Models can be brittle, and apparent precision may hide underlying assumptions or data gaps. For example, an AI-enabled logistics tool might rapidly optimize spare parts distribution based on recent usage rates but fail to anticipate an unexpected operational surge, leading to critical shortages. Although AI improves decision quality and speed within specific functional areas, these gains often remain within silos, each with its own metrics and timelines. Today, AI enables organizations to generate informed options quickly, consistently, and at scale, representing an evolutionary step beyond reliance on individual SMEs.

Functional Silos


(Distinct decision domains within an organization—such as Operations, Maintenance, Training, Logistics, Personnel, and Budget.)

Senior military leaders often describe their role as mastering the art of juggling glass balls, discerning which can be dropped without consequence and which must not fall. This skill, developed through experience, judgment, and expert input, highlights the need to balance competing organizational demands. In practice, these glass balls represent functional silos—distinct decision domains such as Operations, Maintenance, Training, Personnel, Logistics, and Budget, as well as others.

Military organizations are often judged by individual operational success, such as high-stakes missions. However, individual operational success alone does not ensure sustained performance. A commander may execute a mission flawlessly but still weaken the organization if readiness, personnel, or resources are mismanaged. Long-term success depends on synchronizing performance across all silos.

Each functional silo operates by its own standards. Operations focus on mission execution, Maintenance Tracks readiness, Training builds future capability, Personnel manage capacity, and Budget constraints enable them all. Each domain uses its distinct data, metrics, timelines, and definitions of success. While specialization is necessary, it also creates integration challenges.

Success in one silo often creates pressure or risk in another, as resources and time are limited. Decisions that optimize one domain, such as accepting an additional mission, deferring maintenance, delaying training, or reallocating funds, shift risk elsewhere. Because silo outputs are not directly comparable, leaders must reconcile competing signals through experience and judgment.

Senior decision-makers add value by integrating siloed information, weighing trade-offs, and determining which risks to accept to preserve future options. This integration requires both analytical skill and judgment, often relying on trusted advisors and incomplete information. While effective, this approach does not scale easily and is increasingly strained by the speed, complexity, and volume of data and world events.

AI applied within individual silos offers benefits such as improved visibility, faster analysis, and sharper recommendations. However, it does little to address the complex challenge of understanding how decisions in one silo affect others. As long as AI focuses on optimizing siloed performance, senior leaders must still integrate decisions manually.

AI’s true potential lies in helping leaders integrate decisions across silos through a shared evaluative framework. For example, consider a situation where an unexpected mission emerges that requires reallocating resources from Maintenance and Training silos to operational tasks. By using Military Risk as a common language, leaders can systematically assess the potential impact of diverting resources—such as the risk of reduced equipment readiness or diminished future capability—alongside immediate operational benefits. This approach enables leaders to compare trade-offs, align priorities, and manage the organization as a unified system.

Revolution of Military Affairs (RMA)

(A transformative shift in warfare that fundamentally changes how militaries decide, integrate capabilities, and conduct operations.)

Technologies are often considered revolutionary only once their full impact is clear. Rail, telegraphy, precision weapons, nuclear arms, and unmanned systems defined Revolutions in Military Affairs not due to novelty, but because their widespread adoption fundamentally changed military organization, decision-making, and combat. A technology is not revolutionary merely by its existence. Early rail and telegraph systems that provided only localized benefits did not transform warfare. Technologies become revolutionary when they permeate military processes, integrate across functions, and require a fundamental rethinking of decision-making and coordination at scale.

Artificial Intelligence will likely become an RMA-defining technology, though it has not yet done so. Most current defense-related AI accelerates existing processes, such as analysis, output detail, and timelines. While valuable, these improvements often reinforce current decision structures rather than transform them. Senior leaders continue to integrate siloed outputs, arbitrate tradeoffs, and manage risk as before, relying on experience, judgment, and trusted advisors. This underscores where AI’s revolutionary potential remains unrealized.

The true revolution lies not in improving individual decisions, but in changing how decisions are considered collectively. When applied across domains, AI can help leaders compare options, understand cross-functional impacts, and sequence actions over time. In this role, AI does not replace judgment; it enables leaders to view the organization holistically and manage tradeoffs that are difficult to address in isolation.

This is where RMA directly intersects with senior leadership. Revolutionary change does not remove leaders from the decision loop; it changes the loop itself. It compresses time, increases interdependence, and raises the cost of misalignment. In this environment, advantage goes to the organization that best synchronizes information across silos and makes coherent, long-term decisions, rather than simply having the fastest tools.

AI does not need to replace human judgment to be revolutionary. It becomes revolutionary when it enables leaders to manage complexity beyond individual cognition, offering a shared frame of reference for weighing diverse decisions across time and domains in pursuit of a common mission and end state. That shared frame, I argue, is Military Risk.

Military Risk

(The probability and potential consequences of an adverse outcome resulting from military operations, decisions, or resource allocations, assessed in terms of threats, vulnerabilities, and the impact on mission success.)

Senior leaders typically understand the dynamics within individual functional silos. The greater challenge is managing all silos collectively and determining what to protect, defer, or trade off when time, resources, and attention are limited. Success is measured by a unit’s consistent mission accomplishment across all domains while maintaining future capability.

Each functional silo reports performance using domain-specific metrics that often do not translate across the enterprise. For example, budget health may be measured by remaining funds, burn rate, or obligation rates; maintenance by readiness rates, deferred services, or parts availability; and operations by missions accomplished, mission tempo, or capacity utilization. While each metric is valid, they are not directly comparable and may reflect conflicting priorities. Leaders must interpret both the data and its implications for mission success over time and across all silos.

A common language is essential. In this framework, AI must translate silo-specific conditions into their impact on sustained mission success, enabling consistent evaluation of diverse signals. The focus shifts from risk within a single silo to the probability and consequence of those conditions affecting mission achievement now, during critical periods, and in the future. Risk is time-dependent. For example, a training shortfall may pose little risk today but become significant next quarter. Early budget exhaustion may be acceptable if it ensures readiness when most needed.

After translating metrics within each silo, AI can apply weighted decision criteria that reflect the Commander’s intent. These weights adjust as priorities shift, timelines shorten, and critical periods approach. Integrated risk assessments then allow leaders to see how decisions in one domain affect risk in others. This provides a clear view of the trade space: which options reduce overall mission risk, which risks are accepted in specific silos, and how long those risks remain tolerable while pursuing sustained mission success.

The value of this approach lies not in replacing command judgment, but in making its structure visible and defensible. Military Risk becomes a shared framework for understanding functional decisions as parts of a unified system, rather than isolated outcomes. What leaders have long done implicitly, such as balancing priorities, accepting risk, and preserving future options, can now be supported with greater clarity, consistency, and scale. By enabling organizations to make more transparent and structured trade-offs across different domains, this approach fosters organizational coherence and long-term mission effectiveness. This shift, more than any single analytic improvement, highlights AI’s revolutionary potential in military decision-making.

What Follows Integration

My central argument is clear: while applying AI within individual functional silos is valuable, it is primarily evolutionary. The more significant, potentially revolutionary change is integration. By using Military Risk as a shared language, organizations can connect decisions across domains and help leaders sustain mission success over time, rather than achieving isolated successes. This distinction is critical because competitive advantage depends on how well an organization aligns and executes decisions across the enterprise, not on isolated improvements.

The next major shift in AI-enabled military decision-making will occur when AI systems move beyond informing choices. When they can reliably sense the status and activity of functional silos, evaluate options based on time, priorities, and Military Risk, and execute approved decisions through automated systems at machine speed, their role becomes transformative. Some refer to this as agentic AI. At this stage, AI shifts from supporting decisions to providing a true decision advantage.

This future is neither speculative nor distant. Elements already exist in sensing systems, predictive analytics, and automated execution tools. When integrated, these capabilities can significantly compress decision cycles and reduce the cognitive burden on senior leaders. Well-designed systems could help commanders maintain continuous mission success while managing risk across time and domains.

This future also requires caution. Faster decision loops can amplify both advantages and errors. Poor data, fragile models, misaligned incentives, or unclear authority can cause failures at machine speed. The solution is not to avoid progress, but to design it carefully. Importantly, the adoption of faster, more autonomous decision systems raises ethical considerations, such as the responsibility and accountability for decisions made with or by AI. Any move toward greater autonomy must remain grounded in human command responsibility. Guardrails, authority gates, escalation rules, auditability, and deliberate stop points are essential for not only safe and effective speed, but also for ensuring ethical oversight. Commanders must retain responsibility for intent, priorities, risk acceptance, and ethical consequences, even as systems increasingly assist with sensing, analysis, and execution.

These ideas are intended to start a conversation, not end it. If your experience supports, challenges, or complicates this perspective, your input is valuable. I am interested in how others are applying AI in real decision-making environments—what works, what does not, and the questions that arise. This dialogue is essential to advancing our understanding and practice of AI integration. What cross-domain AI experiments have you conducted, and what surprised you? Your field cases could provide invaluable insights as we move forward.

https://www.linkedin.com/pulse/real-ai-advantage-integrating-leadership-decisions-across-zimmer-4lgoe