As you prepare your syllabus for the upcoming semester, a question I’ve explored before looms large. In a previous post, “The Death of the LMS in Higher Ed,” I argued that its role as a simple repository for assignments is becoming obsolete. With AI agents now ubiquitous, how do we prevent the LMS from becoming a mere drop-box for bot-generated work from an autonomous AI Agent?
To show the implications of autonomous AI Agents and the LMS before diving into safeguards and strategies, I wanted to share David Wiley’s (he is an associate professor at Marshall University) video of this in action. Here, he demonstrates how students can use AI agents (OpenAI’s ChatGPT Agent Mode) to infiltrate Canvas and complete assignments and assessments autonomously.
The solution isn’t to fully abandon the LMS (for now), but to evolve its purpose and create humanizing instructional barriers in the tasks and assessments we deliver to our students. This post presents a few practical strategies (I am implementing this term in my own graduate-level course) to revitalize our courses by shifting the focus from the final product, which AI can generate in seconds, to the learning process, which it cannot. This is going to my attempt to defend against students utilizing AI Agents. While I know this may not be one-hundred percent effective, I do have a hunch it will mitigate AI Agents to a degree, at least for the time being, this semester.
Remember, these strategies don’t mean banning AI. The goal is to design assignments and assessments where AI (especially AI Agents) can’t do the heavy cognitive lifting, ensuring the work we assess reflects the student’s own thinking. To do that, we must first be clear on what we’re trying to measure when it comes to student performance and learning. Once this is clear, I will discuss various strategies of what this looks like in secondary high-school classrooms and post-secondary classrooms in colleges and universities.
Performance vs. Learning: What Are We Measuring?
Before designing any assignment and assessment, it’s critical to understand the difference between student performance and student learning.
Performance is the observable, real-time demonstration of a skill, such as correctly answering a question immediately after a lesson. It’s what a student can do right now. While useful for in-the-moment checks, this immediate success can be fleeting and is not a reliable indicator of long-term retention. In practice, this is formative assessment.
Learning, in contrast, is a more durable change in knowledge. It is demonstrated by the ability to reliably retrieve information and apply skills with proficiency over time. True learning means a student can complete tasks independently in the future, proving the knowledge has been successfully encoded into their long-term memory. Our assessment strategies should aim to capture evidence of this durable learning, which needs evidence to support this over the course a longer duration of time.
Strategies for AI-Resistant Assessment
1. Making Thinking Visible with Formative Assessment
Your most powerful tool for capturing in-the-moment performance is the synchronous formative assessment. This strategy can be implemented at the beginning of a lesson, during guided practice, and at the end of a lesson. Using mini-whiteboards or a digital mini-whiteboard (i.e., Pear Deck, Nearpod, Curipod, Wayground, etc.) is highly effective because it demands an immediate, individual response from every student in real-time.
- In a High School Math Classroom: After teaching the point-slope formula, put a problem on the board. Students solve it on their mini-whiteboards and hold them up. You can instantly see who can solve the equation and see who needs support. Do this several times throughout the week as students learn and practice concepts.
- In a University Humanities Course: Ask a complex conceptual question like, “In one sentence, define Foucault’s concept of ‘biopower’.” Students post their answers on a shared digital whiteboard (Pear Deck, Nearpod, Padlet, or Curipod). This gives you a quick snapshot of the class’s understanding in real-time.
2. Flipping the Project Workflow
Reverse the traditional project workflow for any class with a synchronous component. Allow students to conduct initial research at home, but reserve the higher-order work of synthesis and creation for a supervised session during class time.
- In a High School History Classroom: Students research a historical event at home or during independent practice. The in-class assessment involves them working in groups to create a “living timeline” on a large poster as a concept map, defending why they placed events in a particular order.
- In a University Business Course: Students analyze a case study for homework. The synchronous class session becomes a “boardroom simulation” where they present their strategic recommendations and must defend them against critical questions from their peers in breakout rooms. Their responses can be placed directly on a large concept map with their rationale supporting their arguments.
3. Assessing Competency Through Live or Recorded Demonstration
When a skill needs to be demonstrated, require students to show and tell. Narrating one’s thought process while performing a task is a powerful form of assessment. Students can demonstrate live in person or record this through video.
- In a High School Physics Classroom: Students record a short video of themselves performing a procedure, narrating their steps and the scientific reasoning behind them as they explain how a bike pump works to inflate a tire.
- In a University Nursing Program: During a clinical skills check-off (in-person or live online), a student must not only perform a procedure correctly but also explain the “why” behind each action to the evaluating instructor.
4. Documenting the Journey with Iterative Assignments
This strategy is highly effective for all course modalities. Break long-term projects into smaller, sequential assignments within your LMS that document the student’s learning journey, preparing them for a culminating assessment.
- In a High School English Class: An essay assignment is broken down into: 1) An LMS submission of a thesis statement. 2) A peer-review session. The final summative assessment is a timed, in-class or proctored essay for which the prior steps have fully prepared them and synthesized them all into one draft.
- In a University Sociology Course: A research project is scaffolded with due dates for the research question, an annotated bibliography, and a methodology draft. Then, similar to the essay above, students will present their findings live in-person or through a recording or develop a summarized draft of their findings in-class under timed conditions. This process-based work, documented in the LMS, constitutes a significant portion of their grade.
5. Making “Thinking” the Task Itself – Critiquing AI-Generated Content
This is another excellent strategy for any modality. Use AI-generated content as the raw material for an assignment, shifting the focus from content creation to critical evaluation.
- In any High School or University Classroom: Provide students with an AI-generated essay. Their assignment is to act as a critical editor, using the “comments” feature to fact-check the text against course materials, identify logical fallacies, and critique the nuanced arguments the AI missed.
A Note on Online Courses
It’s important to match these strategies to your course modality.
For fully online courses with synchronous components, nearly all of these strategies can be adapted. “In-class” formative assessment checks can be done with digital mini-whiteboards and live polls. Live demonstrations and project work can be conducted in supervised breakout rooms with cameras on. The key is the real-time, synchronous interaction.
For fully asynchronous courses, the options are more limited, but certainly not gone. Strategies requiring real-time supervision are not feasible. However, “Documenting the Journey with Iterative Assignments” and “Assessing Competency Through Live Demonstration” may be better suited for the asynchronous online class environment. These strategies leverage the LMS to track a student’s thought process over time and assess their critical evaluation skills, providing a more robust defense against automated work even without live sessions.
By thoughtfully applying these strategies, we can move beyond the limits of the traditional LMS and design courses that are not only resistant to AI shortcuts but are also more engaging and effective at fostering durable, human-centered learning.
Moving Forward
The key is to be intentional and start small – think less is more. As you review your syllabus for the upcoming semester, I encourage you to select just one major assignment and redesign it using one of these process-focused strategies outlined above. Now, I turn it over to you: What is one strategy you plan to implement this year to navigate and mitigate the use of AI agents and AI-generated content in your classroom? Share your ideas, questions, and challenges in the comments below—let’s learn from each other.
Navigating the Present and Future of Education 