Literacy in Artificial Intelligence

The Need for Creating Literacy in AI and Large Language Models in Army Professional Military Education

By COL Emiliano Tellado

June 29, 2026 | Field Artillery, 2026 E-Edition

Read Time: < 14 mins

A rtificial intelligence (AI) is reshaping the character of war, accelerating decision cycles, and transforming how militaries plan, fight and sustain operations. AI-enabled systems such as automated target recognition, sensor-fusion platforms, and predictive analytics are already influencing how militaries sense, understand and act on the battlefield. At the same time, large language models (LLMs)—such as those behind Microsoft Copilot, ChatGPT, Gemini and Claude—have become embedded in target development, knowledge management and staff work. Yet across the Army, leaders’ understanding of these technologies remains uneven, often shaped by misconceptions or limited exposure. As the Army modernizes, the gap between technological capability and human comprehension poses a risk. The force cannot responsibly employ emerging tools if its leaders do not understand what they are, how they work and where their limits lie. If the Army is going to meet the goals for AI-driven command and control at the theater, corps and division headquarters by 2027, the time is now for action.1

This challenge is fundamentally a literacy challenge. “AI Literacy is the ability to comprehend various aspects of artificial intelligence, including its capabilities, limitations, and ethical considerations, and to use it for practical purposes. It might entail learners exercising critical thinking in their understanding of AI technologies and their applications.”2 LLM literacy is the ability to use language-based AI tools effectively, evaluate their outputs and recognize when they support or undermine sound judgment. Literacy is not a technical specialization; it is the conceptual foundation leaders need to integrate emerging tools into their workflows responsibly.

This article argues that AI and LLM literacy must become a core leadership competency and be deliberately integrated into Professional Military Education (PME). It outlines what Soldiers currently understand and misunderstand about AI and LLMs, clarifies the distinction between them, identifies practical applications that improve efficiency across the force and proposes a gated PME framework that builds literacy progressively across all cohorts.

The View from the Force: What Soldiers Know

To understand why AI and LLM literacy must become a PME priority, we must begin with the force itself. In focus groups with young officers, NCOs, warrant officers, and enlisted Soldiers, a clear pattern emerged: Despite daily exposure to AI-enabled tools in civilian life, most participants had little understanding of what AI is, how it works or how it could support their military roles. Many equated AI with chatbots or automation, and others assumed AI systems “think” or “understand” in human terms. A significant number expressed discomfort or avoided AI tools entirely because they lacked the knowledge to use them. “Pew found that one in 10 students uses chatbots to do most or all of their schoolwork. And nearly 60% of teens believe that their peers are regularly using AI to cheat at school.”3 Schools are still struggling to preserve the merits of creativity and ingenuity while leveraging tools such as AI and LLMs.

These discussions highlighted a deeper issue: The Army is fielding increasingly sophisticated technologies to a force that lacks the conceptual foundation to employ them responsibly. ADP 3-13 acknowledges that “To improve data literacy across the force, the Army employs data literacy programs of instruction across the Army’s education system.”4 However, implementation across PME remains uneven, with most institutions still in the early stages of integrating structured AI and data-literacy instruction. Soldiers are not resistant to AI; they are unprepared for it. Many are unfamiliar with how to use AI even for basic tasks such as drafting a letter or reviewing a document.

The real gap is operational judgment: knowing when to use AI or an LLM, how to use it effectively, and where it fits within a workflow or decision process. Soldiers repeatedly asked:

• When is an LLM helpful, and when can it introduce risk?

• How should a task be structured so AI can support it?

• Where does AI fit within Military Decision-Making Process (MDMP), training development or maintenance management?

The issue is not resistance but experience; leaders need the conceptual foundation to employ these tools responsibly. Without a mental model for how these tools function, Soldiers cannot integrate them into their workflows with confidence or precision.

AI vs. LLMs: A Critical Distinction

Clarifying the distinction between AI and LLMs is only the first step. Once leaders understand what these tools are and what they are not, they must learn how to employ them responsibly. Literacy bridges this gap: It transforms conceptual understanding into disciplined application. The next challenge is knowing when and how to automate tasks without surrendering judgment. These gaps reveal a foundational problem: The force lacks a clear conceptual map of what AI is and is not. Before the Army can teach responsible use, it must correct this confusion. A central barrier to responsible adoption is conceptual confusion: Many Soldiers use AI to describe the LLM-based tools they encounter most often, but AI and LLMs are not interchangeable.

AI is the broad field of computational methods that enable machines to perform tasks requiring human-like cognition, such as pattern recognition, prediction, classification, perception and planning. It includes machine learning, robotics, computer vision and autonomous systems. These capabilities already underpin predictive maintenance, sensor fusion and emerging autonomous platforms. LLMs, by contrast, are a specific type of AI trained on massive text datasets to generate and interpret human language. They excel at drafting, summarizing, and synthesizing information, but they do not “think,” “reason” or “understand” in the human sense. LLMs’ strengths—speed, fluency and breadth—make them powerful accelerators of staff work, while their weaknesses—hallucinations, overconfidence and sensitivity to input phrasing—make them unreliable as decision-makers.

Treating LLMs as general-purpose AI obscures these distinctions and invites misuse. PME must correct this by giving leaders a clear conceptual map of the broader AI ecosystem and the specific role LLMs play within it.

Understand First, Then Automate

Once leaders understand the tools, the next challenge is understanding how to employ them responsibly. That begins with a foundational principle: Leaders must understand a task before they automate it.

One of the most important lessons emerging from both research and field experience is simple: Leaders must understand a task before they automate it. “LLMs can generate inconsistent, hallucinated, or biased responses, and the same prompt can yield different outputs.”5 Leaders who do not understand the subject, the task or the tool’s settings cannot judge whether an AI-generated output is accurate or appropriate. As Santiago notes, “No matter how much information you have or how advanced your tools are, your data will not tell you much if you do not start with the right questions.”6 Literacy is what enables leaders to ask those questions, verify outputs, and apply human judgment. AI is not a substitute for expertise, but it can enable leaders to make more informed decisions if it’s used with an understanding of the subject and the information it uses to reach its conclusions. This principle applies across the force:

• A staff officer who does not understand MDMP cannot rely on an LLM to produce a mission-analysis brief.

• A platoon leader who does not understand maintenance trends cannot rely on predictive analytics to identify readiness risks.

• A trainer who does not understand learning objectives cannot rely on AI to build effective lesson plans.

AI is not a substitute for expertise. It is a tool for experts to automate tasks and reduce redundant work. PME must reinforce that responsible use begins with understanding how these systems think, not assuming they do. If leaders do not understand the limitations of LLMs, such as how they interpret data and how they receive prompts that shape the answers that they provide, they may fall into a trap of overconfidence or disillusionment with the tools early on. “Researchers found that models can mistakenly link certain sentence patterns to specific topics, so an LLM might give a convincing answer by recognizing familiar phrasing instead of understanding the question.”7 These risks do not justify avoiding the technology; they underscore the need for leaders who can work with it confidently and demand from industry partners the tools that reduce these vulnerabilities. Commanders who understand these tools will shape their formations’ ability to think and integrate these capabilities as they arrive, creating a space for innovation.

The current generation of AI tools—such as LLMs, predictive analytics and automated processing systems—represents only the first wave of capabilities Soldiers will encounter. Emerging technologies such as edge-based physical AI, autonomous robotic systems and increasingly capable multimodal models will further accelerate decision cycles and push more cognitive tasks to the edge. These systems will not replace leaders, but they will demand leaders who understand how AI interprets data, how prompts shape outputs and where algorithmic confidence can mask underlying errors. Literacy built today becomes the foundation for integrating tomorrow’s capabilities, ensuring that the Army does not merely catch up to current technology but seizes and holds the initiative as more advanced AI systems enter the force.

Practical, Everyday Uses of AI & LLMs in the Army

LLMs can provide benefits across the force when used with subject-matter understanding and within OPSEC, classification and policy constraints. They help Soldiers accelerate research, refine written products and generate initial drafts of routine documents. These tools support expertise and streamline routine work, allowing leaders to focus on higher-order thinking and decision making. Across warfighting functions, leaders can use LLMs to accelerate Intelligence Preparation of the Battlefield (IPB), refine running estimates, draft initial targeting products, or consolidate doctrinal references during MDMP, always with human verification.

Proofread, edit and clarify intent.

LLMs can help Soldiers refine emails, orders, briefings and academic writing. More importantly, they can help leaders verify that their intent is being communicated clearly.

For example, a leader drafting a memorandum can ask an LLM:

• Is my intent clear?

• Does this paragraph read as a directive or advisory?

• Is the tone appropriate for a senior audience?

Draft initial products when the Soldier already understands the task.

The Army must acknowledge that exposure does not equate to experience. While the current generation of young leaders has been exposed to this technology at an earlier age, it does not mean they understand how to use it. Therefore, we must highlight practical uses to build the necessary understanding.

Consider a platoon sergeant preparing a quarterly counseling statement. They know what they want to say but are unsure about tone and structure. An LLM can help by:

• Suggesting a professional, balanced tone

• Offering sample phrasing aligned with Army counseling standards

• Helping articulate expectations clearly

Or consider a captain refining a memorandum for a battalion commander. The captain can ask an LLM:

• Does this memo clearly express the decision I’m recommending?

• Is the tone appropriate for a field-grade audience?

• Where might a senior leader misinterpret my intent?

PME must formalize the literacy that provides Soldiers with the familiarity needed to make these tools reliable, responsible and effective.

Integrating AI and LLM Literacy into PME: A Gated, Iterative Approach

These examples show what Soldiers can do today, but they also reveal why structured education is essential. To prepare leaders for a future shaped by emerging technologies, PME must adopt a gated, iterative approach.

If the Army expects leaders to employ AI and LLM tools responsibly, PME must provide a structured and progressive learning pathway. This begins with a simple question: What do we need Soldiers and leaders to understand, and at what point in their professional development should they understand it? PME already uses a gated model—Basic Officer Leadership Course (BOLC), Captain’s Career Course (CCC), Command and General Staff College (CGSC), and the Army War College (AWC) for officers; Advanced Individual Training (AIT) and NCO PME for enlisted and NCOs; and a parallel Warrant Officer Education System (WOES). AI and LLM literacy must follow the Army’s existing gated-education model with each PME level reinforcing and expanding the previous one.

Establishing Baselines: What Leaders Need to Understand at Each Gate

The first step is defining the baseline knowledge required at each level of PME. These baselines should answer three questions:

• What terms and concepts must leaders understand at this level?

• What open-source or Army-provided tools should they be able to use?

• What tasks should they be able to perform with AI or LLM assistance?

This ensures that Soldiers do not encounter AI literacy as a one-time event but rather as a progressive skill set that grows with responsibility.

Iterative Development Across PME

AI literacy must be iterative, with each PME gate reinforcing and expanding the previous one. This aligns with the existing PME design.

Officer Pathway

• BOLC — Foundation

  • Introduce core terms: AI, LLMs, machine learning and data quality

  • Teach responsible use of open-source tools

  • Practice simple applications: summarizing doctrine and drafting counseling statements

  • Build confidence and familiarity

• CCC — Application

  • Integrate AI and LLMs into staff processes: MDMP support, research consolidation, information synthesis

  • Teach clarity checks for written products

  • Introduce AI-enabled systems relevant to branch functions

• CGSC — Integration

  • Apply AI tools to operational planning, wargaming, intelligence preparation and sustainment analysis

  • Teach leaders how to evaluate AI-enabled systems and

    interpret outputs

  • Introduce ethical, legal and policy considerations

• AWC — Organizational leadership

  • Prepare leaders to shape AI adoption across formations

  • Evaluate policy, risk and organizational processes

  • Mentor subordinate leaders on responsible use

  • Integrate AI into strategic planning and modernization efforts

Warrant Officer Pathway

Warrant officers occupy a unique space: technical experts, system integrators and advisors. Their AI literacy pathway must reflect that.

• WOCS — Foundation

  • Introduce AI and LLM fundamentals

  • Teach responsible use and MOS-specific applications

• WOBC — Technical application

  • Integrate AI tools into MOS-specific systems (e.g., fires, aviation, intelligence, maintenance)

  • Teach LLM-supported troubleshooting, data interpretation and documentation

• WOAC — System integration

  • Apply AI to unit-level processes: readiness analysis, maintenance forecasting and data management

  • Teach warrant officers to advise commanders on tool selection and employment

• WOSC/WOSSC — Technical leadership

  • Prepare senior warrants to shape organizational AI adoption

  • Evaluate system performance, data integrity and integration challenges

  • Mentor junior warrants and staff on responsible use

  • Warrant officers become the connective tissue between emerging technology and operational employment, making their literacy essential.

NCO and Enlisted pathway

• AIT — Introduction

  • Teach basic terminology and responsible use

  • Introduce simple LLM applications tied to MOS tasks

• BLC/ALC — Practical application

  • Draft counseling statements, training calendars and SOP updates

  • Use LLMs to summarize regulations or refine written products

  • Introduce AI-enabled tools relevant to MOS functions

• SLC/MLC — Unit-level integration

  • Apply AI to training management, maintenance processes and staff support

  • Teach NCOs how to mentor Soldiers in the responsible use of technology

• SGM Academy — Organizational influence

  • Prepare senior NCOs to advise commanders on AI integration

  • Manage organizational adoption and training

  • Shape unit-level policy and culture

Across all cohorts, AI and LLM literacy should progress through three themes—understanding terms and tools, applying them to real tasks and introducing advanced functions—and each PME gate builds on these themes at the appropriate level.

Figure 1. Leading the Machine: Building AI & LLM Literacy in the Army 8 The Army is rapidly adopting AI-enabled systems, yet a significant literacy gap exists among leaders. To ensure responsible and effective employment of these tools by 2027, the Army must move beyond “exposure” and integrate a progressive, gated education model across all cohorts.

A Scenario That Brings It All Together

Imagine a captain at CCC preparing a mission analysis brief. They use an LLM to consolidate doctrinal references and generate an initial list of specified and implied tasks, then manually verify each one. When the captain inputs the mission variables, the LLM highlights a potential sustainment friction point tied to terrain and route constraints, something the staff had not yet discussed. The captain validates the insight, incorporates it into the brief and the commander adjusts the COA accordingly. AI literacy built at BOLC and reinforced at CCC enables the captain to work faster and identify operational risk earlier.

Here’s how AI literacy taught progressively across PME helps them:

1. They use an LLM to summarize key doctrinal references, saving an hour of reading time.

2. They ask the LLM to identify potential friction points based on the mission variables they provide.

3. They independently verify the key points, ensuring accuracy and relevance.

4. They use the LLM to refine the tone and clarity of their slides, ensuring their intent is unmistakable.

5. They integrate AI-enabled systems (e.g., predictive maintenance data or ISR summaries) to inform their assessment.

AI literacy built at BOLC, reinforced at CCC, expanded at CGSC, and refined at AWC enables Soldiers to work faster, think more deeply and present more clearly.

Conclusion

AI and LLM literacy are now a prerequisite for effective leadership in a modernizing Army. The force is already encountering AI-enabled systems on the battlefield and LLM-based tools in daily staff work, yet focus groups show that Soldiers lack the conceptual foundation to employ them responsibly. This gap threatens the Army’s ability to achieve the decision-advantage goals embedded in modernization and data-centric command and control. “The side that is best prepared, best understands an operational environment, adapts more rapidly, and acts more quickly in conditions of uncertainty is the one most likely to win.”9

PME must close this gap by giving leaders a shared vocabulary, a clear understanding of what AI and LLMs can and cannot do and practical experience applying these tools to real tasks. Human judgment must remain the final authority in all decision-making, but judgment is only as strong as the literacy that informs it.

The Army cannot afford a future where its tools outpace its people. Building AI and LLM literacy now ensures the force remains ready, adaptive, and grounded in the enduring principle that technology serves the commander, not the other way around. The leaders who master this literacy will shape the Army’s modernization; those who ignore it will be overtaken by it.

References