Artificial intelligence is revolutionizing assessment practices in military education and in the training environment, offering unprecedented efficiency and consistency while preparing future leaders for an increasingly AI-driven operational landscape.
At the US Army War College, we recognize that embracing AI tools enables instructors to focus more deeply on assessing students' developing critical thinking and strategic analysis skills that define exceptional military leadership. For leaders, understanding and practicing with AI ensures relevance in one of the fastest growing technology leaps in a generation.
Our approach maintains the delicate balance between technological advancement and human judgment. By implementing AI integration strategically, we uphold our commitment to educational excellence while adapting to emerging technological realities.
By pushing the limits of AI, you discover its capabilities and constraints. Only then can you craft a blueprint for students, or soldiers, by showcasing your AI journey, igniting a spark for innovation.
Considerations for AI-Assisted Feedback
Using Large language models (LLMs) to speed feedback and refocus faculty on content.
Basic AI Concepts
Large language models (LLMs) are advanced computer models designed to understand and generate human-like text. They're trained on vast amounts of text data to learn patterns, language structures, and relationships between words and sentences.
Identify language patterns matching rubric criteria
Flag missing sections or terminology
Scan for grammatical errors and inconsistencies
Draft concise or detailed feedback
Identify themes and patterns across multiple submissions
Limitations of AI Integration in Feedback
Requires explicit prompts for deep insight
Risk of "hallucinating" supporting evidence
Potential bias towards certain writing styles
Struggles with subtle reasoning without exemplars
Difficulty grading true analysis over restatement or excessive quoting
In the example of integration AI into a feedback session or product, the rubric and faculty/leader assessment remain the ultimate authority, not the AI's interpretation. Always reference exact criterion wording in your prompts. AI enhances and accelerates your application of the information; it does not replace human judgment.
How LLMs Work: Predicting the Next Word
Pre-trained Data Sets
LLMs are trained on massive amounts of text data (books, websites, articles). They learn patterns in how words and phrases typically appear together.
Mathematical Predictions
At their core, LLMs use complex mathematical models (neural networks) to calculate probabilities. They predict "what word is most likely to come next" based on the patterns they learned.
Language Syntax & Context
The model analyzes the syntax (grammar structure) and context of your input to determine which words statistically fit best. It's not "thinking" - it's calculating probabilities based on patterns.
Token by Token Generation
LLMs generate responses one token (word or word piece) at a time, each prediction building on the previous ones. Each word is chosen based on what's most probable given everything that came before.
LLMs don't 'understand' meaning - they're incredibly sophisticated pattern-matching systems that predict likely word sequences based on statistical relationships in their training data.
Levels of AI Integration for Leaders and Educators
Quick Guide to Modules
Navigate the essential topics covered in this comprehensive guide.
See how AI acts as a force multiplier in military education.
Important for Faculty: As part of the prompting process for all of these modules, upload your organization's writing guidance, the course rubric, and any relevant course content as PDFs so the LLM can anchor its responses to official standards and language.
AI as a Tool
Ways to Use AI
Use embedded AI tools (ie. Grammarly) for grammar checks, citation verification, and plagiarism detection.
Use AI to provide examples of common writing errors (see below) or provide suggestions to fix errors.
Have AI create a variety of topics, flash card, or prompts to expand lessons/professional development sessions.
PRO TIP: Run your collective feedback/comments through AI to find themes and commonality for a particular class or work center
Faculty/Leader Role
Review AI suggestions, accept useful inputs while discarding inappropriate ones based on context.
Speeds administrative work (lesson building, MFR writing) by having preset language easily adapted to individual context. For faculty, preset questions or writing corrections replaces need to create new ones for each assignment.
Faculty/leaders now focus on providing personalized, substantive feedback or comments that AI cannot provide, especially regarding strategic thinking and military operations.
Considerations
Most individuals will likely use Grammarly and other grammar tools for their writing. Highlight errors appropriately and provide clear examples of correct verbiage to facilitate learning.
This level of AI usage leverages platform-embedded AI for structural review and LLMs to create positive writing examples. No student works or identifiable information are uploaded to AI models.
Different AI models provide varying levels of response quality. The comparison below is a simplistic example to showcase Gemini Pro against Microsoft Copilot using the same starting prompt "You are a professional military faculty member giving feedback on common writing mistakes such as passive voice, informal voice, subject-verb agreement, comma splices, run-on sentences, and pronoun disagreement. Other common pitfalls include wordiness, using the wrong words (especially homophones). Provide examples of the errors using military topics along with corrected versions for comparison."
Create feedback statements using predefined criteria.
Create AI responses to the assignment aligned with the rubric to serve as examples or guides.
Feed AI generic personality profiles (i.e., IFTJ, ENTP) and have it create more customized language based on your feedback.
Have AI summarize an image of group whiteboard work.
PRO TIP: Using AI as another "student" provides a neutral platform for students to practice providing critical feedback and learning the limitations of AI in the process.
Faculty/Leader Role
Faculty use AI-generated statements during grading to provide rubric-aligned and/or personality-based feedback, saving time.
Using AI to capture group work from white-board or game sheets allows more insight by identifying patterns and thinking processes. These summaries are useful in future grading or operational planning.
Faculty/Leaders still focus on providing substantive feedback that AI cannot provide.
This level of AI usage provides criteria-grounded feedback examples to speed grading, planning, or as an innovative teaching/discussion tool. No student works are "uploaded" to AI models. AI integration increases.
The "Perfect Prompt for Rubric Grounded Feedback"
File upload
Prompt for rubric grounded feedback.docx
19.6 KB
Rubric Example from Gemini
Gemini
‎Gemini - direct access to Google AI
Created with Gemini
Rubric Example from ChatGPT
ChatGPT
Grading rubric translation
A conversational AI system that listens, learns, and challenges
Use AI to create a unknown link for differing learning styles (audio, video, flash cards, etc).
Have AI generate updated lesson plans, scenario examples, or create images to enhance program.
Create an AI agent to function in varying assistant roles (wargaming , Socratic tutor, operational planning).
unknown link: "I'm having a persistent problem with [insert topic] in [insert job or department] despite having taken all the necessary countermeasures I could think of. Ask me enough questions about the problem to find a new approach."
Faculty/Leader Role
Military educators/leaders still lead the process of proving in-line comments on writing and critical thinking, and of being the final decision maker on planning products. AI is not used to determine a grade or force a choice.
Faculty and leaders speed the feedback process by submitting deidentified assignments/products to an AI to complete "first-pass" feedback and identify immediate issues.
Updating lesson materials quickly allows faculty's more time for scholarship or student interaction.
AI agents are customized conversations geared toward tasks that leaders train as needed. These agents can act as a substitute for missing planning members or serve as a red team.
OPSEC Considerations Use only approved AI platforms. Do not upload CUI, PII, sensitive or classified information to uncleared AI systems. Ensure you anonymize or de-identify all personal information before uploading assignments into AI.
Deploy AI Agents capable of workload without human intervention.
Course themes and feedback
Advanced Student Feedback
Big Data analytics
Presentation creation (*this presentation was made with AI assistance in Gamma)
Brainstorming: exploring research ideas, expanding napkin conversations
Faculty/Leader Role
For this level of use, faculty/leaders need a robust understanding of AI tools, and platforms. They should have enough experience with AI to recognize and address limitations.
Ethical imperative: Maintain faculty review for high-stakes assessments like capstone projects, strategic analyses, and leadership evaluations.
Examples of AI Partners
NotebookLM allows faculty and students to organize their research or lesson materials for easier distillation, summarization, search, and overall engagement with sources. Try this notebook: Curly's China Course and then use the AI Gem to evaluate student material for a particular week.
Military-specific language, doctrinal frameworks, and strategic concepts may be misinterpreted by AI systems trained primarily on limited or unclassified data.
Solution: Regularly audit AI outputs for systematic bias; supplement general AI models with military-specific training data; implement faculty override protocols.
Privacy and Security
Military and student work often contains sensitive strategic analyses, references to classified materials, or discussions of current military operations.
Solution: Use only approved AI systems, follow your institutional policy. Here is USAWC policy.
File upload
AY26 USAWC Generative AI Policy-1.pdf
AY26 USAWC GAI Policy
Academic Integrity
As AI writing tools evolve, distinguishing between student-generated and AI-generated work becomes increasingly challenging.
Solution: Combine AI detection tools with honor code reinforcement; design assessments requiring in-person components; develop authentic military scenario-based evaluations that require genuine expertise.
Begin with one platform and less complex tasks before applying AI to high-stakes workflow.
Get comfortable with AI and expand your platform use.
Promote Transparency
Clearly communicate with students and soldiers about when and how AI is used.
Provide opportunities for students and soldiers to incorporate AI when appropriate.
Develop AI Literacy
Participate in regular training on AI capabilities, limitations, and ethical considerations specific to military education and operations.
Collaborate with colleagues to share best practices and lessons learned from AI implementation.
Continuous Evaluation
Regularly assess AI accuracy against benchmarks to ensure alignment with standards.
Document and address any systematic discrepancies between AI and human assessment patterns.
Development Opportunity: Host a workshop on AI integration, join an ongoing development session on AI, take an online course in AI, or simply practice!
Embracing the Future: AI as a Force Multiplier in Military Education
"Strength and Wisdom"
Our motto guides us as we harness AI to enhance learning, uphold integrity, and prepare leaders for tomorrow's challenges.
Our Commitment
The US Army War College remains dedicated to balancing technological innovation with the irreplaceable human elements of military education. As we integrate AI tools, we strengthen our ability to develop strategic leaders who think critically, communicate effectively, and lead with integrity in complex environments.
Become a pioneer in ethical AI use by participating in our AI Acceleration Committee. Share your expertise, concerns, and insights to help shape the future of military education in the digital age. This is not just a "fun" thing to do, it is a national security imperative.