Creating Assignments that Eliminate the Use of AI
Large Language Models (LLMs) like ChatGPT are increasingly sophisticated, and their output can be difficult to distinguish from student-generated work. Because AI detection tools are unreliable and prone to false positives or negatives, relying on them is not a sustainable solution. Instead, instructors should focus on assignment design strategies that reduce the incentive or utility of using AI in the first place. While no assignment is completely “AI-proof,” thoughtful design can make AI use irrelevant, impractical, or counterproductive.
Interesting in incorporating AI into a writing assignment-- check out our companion article!
When revising your assignments in light of the rise of AI tools, consider these reflection questions borrowed from Derek Bruff’s article “Assignment Makeovers in the AI Age”:
Why does this assignment make sense for this course?
What are the specific learning objectives for this assignment?
How might students use AI tools while working on this assignment?
How might AI undercut the goals of this assignment? How could you mitigate this?
How might AI enhance the assignment? Where would students need help figuring that out?
Focus on the process. How could you make the assignment more meaningful for students or support them more in the work?
Below are several effective approaches that emphasize critical thinking, incorporate personalization as a student-centered practice, and focus on the learning process, not just the product.
Scaffolded Assignment Design
The easiest way for students to use AI is to ask for a completed essay responding to a single given prompt. Thus, requiring multiple deliverables makes going straight to an AI-generated solution more difficult. Break larger writing projects into smaller, sequential tasks with clear instructions, deadlines, and feedback checkpoints. This encourages student engagement with the process of writing—not just the final product.
For example, you might require:
A proposal or brainstorming document
An annotated bibliography or research log
A thesis statement with an outline
A rough draft with peer feedback
A revision memo explaining changes
Students are less likely to use AI when each stage requires reflection, feedback, and revision. This approach also models academic writing as a process, not a one-time product.
Reflective and Metacognitive Activities:
Incorporating reflective prompts encourages students to draw on personal experiences and develop self-awareness about their learning process. These are difficult for AI to fake persuasively and foster student ownership of their work.
Ask students to provide a self-assessment of their work, reflect on their specific learning journey, or describe how they solved a problem:
“What part of this assignment did you find most challenging, and how did you approach it?”
“What feedback from peers or the instructor helped you the most?”
“How has your thinking changed about this topic over the course of the quarter?”
On exams or discussion boards, consider including prompts like:
“Choose a homework problem you struggled with this semester. What did you learn from the process of grappling with it?”
“Which reading most changed your understanding of the course material, and why?”
“Which class discussion or peer comment helped you see something in a new way?”
These types of questions not only deepen learning but make AI-generated answers much more obvious and unconvincing.
You might also tailor your prompts to individual students or class discussions. AI has no memory of the student's earlier work. Personalized tasks ask students to reflect on or extend ideas they've already shared, making it harder to rely on generic or pre-written responses while supporting student-centred learning. Personalization encourages deeper engagement with course material and affirms students’ unique' voices and perspectives. It supports student-centered learning, helping students make meaningful connections between new knowledge and their own experiences.
Consider asking students to:
Revise and expand on a previous discussion post in light of new material
Build a project around a topic they proposed earlier in the quarter
Connect course content to their major, work experience, or future goals
Respond to instructor comments on a draft with a targeted revision plan
This approach not only limits AI’s usefulness but also strengthens the coherence and relevance of students' learning across the term.
Applied Learning & Collaboration:
Ask students to apply course concepts to specific, real-world situations, particularly local or recent events. This situational specificity makes it harder for AI to generate relevant, accurate, or nuanced responses.
Examples include:
Connecting a theory to a current news story from your region
Analyzing a local policy or organization through the lens of course material
Interviewing a community member or writing a profile that ties back to class themes
The more situated and personal the task, the harder it is to offload onto AI. Moreover, students are able to connect what they learn in the classroom more meaningfully to their everyday experiences.
You might also foreground collaborative assignments. Design team-based assignments where students must negotiate roles, responsibilities, and shared outputs. AI struggles with interpersonal dynamics, evolving dialogue, and negotiated meaning. When students are responsible for different aspects of a group project (e.g., research, synthesis, presentation, peer feedback), they are accountable not only to the instructor but to each other. Collaboration promotes active learning and interdependence, and it also provides multiple opportunities for in-process checkpoints, making it more difficult to rely on AI-generated work without being noticed by group members.
Examples include:
Group research projects that require division of labor and a shared annotated bibliography
Collaborative writing assignments using tools like Google Docs with visible revision history
Peer review or workshop days where students critique and refine each other’s work
Structured debate preparation or case study analysis in rotating teams
By embedding collaboration into the learning process, instructors make writing and thinking public and dialogic—something AI simply can’t replicate convincingly.
Beyond AI Deterrence: Better Teaching Practices
Ultimately, the strategies outlined above aren’t only about preventing academic dishonesty; they’re about promoting meaningful learning. Assignments that emphasize reflection, personalization, collaboration, and real-world application foster deeper engagement, stronger critical thinking, and more transferrable skills.
The rise of AI in the classroom challenges us to reexamine our pedagogical practices. But rather than framing this moment solely as a threat, we can see it as an opportunity: a chance to make our assignments more human, more authentic, and more aligned with the kinds of thinking we want students to carry beyond our classrooms.