The Butler vs The Sparring Partner: Reframing the AI Relationship for Students
How changing one metaphor can transform AI from academic threat to learning catalyst
Here's why it's so hard to change how students use AI in the classroom: Students don't just randomly decide to use AI as a shortcut. They're responding to every signal they receive about what AI is for.
What Students See When They Open AI
When students (and users in general) open up AI, this is what they see:
A butler, a servant, a personal assistant who will do anything for you — just ask!
The interface design, the marketing language, even the name "assistant" all reinforce one clear message: AI is here to serve you. AI exists to make your life easier by doing work for you.
This butler metaphor feels intuitive to students because it aligns with how they've learned to think about technology in general. Tools should be convenient. They should save time. They should solve problems.
But this default assumption is preventing students from engaging with AI in ways that might actually benefit their learning and skill development.
The Reframing We Need
In order to develop AI literacy in students and users broadly, we need to reframe the relationship. We need to teach users to engage with AI as a sparring partner, not a servant.
Learning to see AI as a sparring partner who challenges their thinking, tests their knowledge, and makes them stronger through the interaction. The difference isn't just about preserving academic integrity. It's also about whether AI becomes a lifelong crutch that weakens student thinking or a catalyst that strengthens it.
When students treat AI as their academic butler, they're missing the most powerful opportunity AI presents for learning - and also opportunities for educators to evaluate process-based thinking on the page.
The Sparring Partner Alternative
This shift in metaphor changes everything about how students (and users in general) engage with AI.
When AI is a butler, the question becomes: "How can I get this done faster?"
When AI is a sparring partner, the question becomes: "How can I use this interaction to demonstrate and improve my skills?"
What Changes When Students Spar with AI
The Work Becomes Visible
With the butler approach, students try to hide their AI use. The goal is for AI's work to be invisible, seamlessly integrated into their final product.
With the sparring partner approach, the AI interaction becomes part of the learning process that educators can see and assess. For over a year, I have been trying to prove that it is possible to grade student interactions with AI. This, I realize now, is the first step. If students use AI as a butler at any point in the learning or assessment, the educator will have nothing to grade. Or, interactions will be hollow, leaving little opportunity for meaningful feedback.
But within this model, students aren't trying to hide their AI use, they're demonstrating their ability to engage with it thoughtfully. It’s metacognition on the page - an opportunity for direct metacognition instruction unlike anything we have ever had before.
Skills Get Demonstrated, Not Delegated
Consider these two approaches to the same writing assignment:
Butler Approach: "ChatGPT, write me a 500-word essay about climate change."
Sparring Partner Approach: "I'm arguing that climate policies should prioritize adaptation over mitigation. Help me identify the strongest counterarguments to this position so I can address them effectively."
In the first scenario, the student delegates the thinking. In the second, they demonstrate their understanding of the topic, their ability to construct an argument, and their willingness to engage with opposing views.
The Subject Matter Expertise Test
Become a paid subscriber for in-depth content focused on embedding meaningful AI Literacy into the classroom. You’ll also get advance access to toolkits, lesson plans, and more that can be used across a wide range of educational settings.
Another powerful aspect of the sparring partner approach is how it reveals genuine subject matter expertise.
When students ask AI to write or produce around topics they don't understand, the limitation becomes obvious quickly - and lives directly inside of the interaction. They can't evaluate AI's responses, can't ask follow-up questions, and can't direct the conversation toward deeper insights.
But when students have real knowledge about a subject, their AI interactions become sophisticated. They can spot errors, build on AI's ideas, and push the conversation into nuanced territory that reveals their expertise. Thus, if you have already established a certain level of expectations with respect to the performance of subject matter expertise, you can also see this on the page.
This is why the ‘backboard’ or sparring partner approach works so well in academic settings. It doesn't eliminate the need for subject matter knowledge - it makes that knowledge more visible and more valuable.
Practical Implementation for Educators
So how do you help students make this shift? Here are three strategies that work:
1. Change Your Assignment Language
Instead of: "Write a research paper on..." Try: "Use AI to help you stress-test your argument about..."
Use examples like the below simple re-framing prompt to help students understand the best use of AI — one that protects their cognitive development and creativity, preserves assessment integrity, and prepares them for an AI world.
Instead of: "Don't use AI for this assignment" Try: "Include a reflection on how your AI interactions revealed or challenged your understanding of..."
My very first assignment with AI required a reflection. For whatever reason, it seemed to get buried beneath the concept of “grading the chats.” But in that assignment, my students had to analyze their own approach to interviewing a Holden Caulfield chatbot and subsequently share:
A) How the bot helped them understand the character (if at all)
B) When and where the interaction was useful (if at all)
C) How their perception of AI and their relationship to it changed as a result of the interaction, and
D) Whether or not character bots should be considered a useful tool for learning more about a subject, character, or topic.
**The key when implementing AI reflections is to recognize that there is no right answer to any of these questions. Every interaction is different, and the final outcomes are all subjective. Instead, the reflections become places for increased awareness of both self and tool - as well as opportunities for students to provide language-based justifications of their own opinions.**
Below is the initial broad framework I published in support of this goal:
2. Make AI Interaction Part of the Assessment
Ask students to submit their AI conversation along with their final work. Evaluate not just the final product, but the quality of their AI engagement. Did they ask thoughtful questions? Did they build on AI's responses? Did they catch errors or push back when AI missed something important?
One crucial nuance is that this process becomes significantly more effective when students are given clear benchmarks for a meaningful sparring partner interaction. For more on this subject, I encourage you to read my Foreword for The Field Guide to Effective GenAI Use in The WAC Repository.
**Note to Reader: I see this method as a once-a-semester experience, especially in the early days as we gather data on what works and what doesn’t in the context of our traditional academic goals. I do not see this as an activity for every day, every week, or every month, which is a common misconception.**
3. Model Sparring Partner Interactions
Show students what effective AI engagement looks like. Demonstrate how to ask follow-up questions, how to challenge AI's responses, and how to use AI to explore ideas rather than generate finished products.
Again, comparison is absolutely crucial. It’s not enough to give them a single prompt and expect the entire interaction to follow this mold. As I wrote in the aforementioned Foreword of The Field Guide, providing mentor prompts and expecting a thoughtful AI interaction is like a Writing Instructor giving students sentence stems and expecting a high-quality essay.
A Note on Attribution and Community
I first developed and wrote about the concept of setting up “adversarial” interactions with AI in the spring of 2024 - in conjunction with Ryan Tannenbaum. Since then, educators across the country have picked up and/or developed these ideas or arrived at them independently, which is fantastic! Not only do I support this organic spread of ideas, but I actively encourage it.
However, if you find yourself changing your assessment protocols or reconceptualizing AI in your classroom because of this work, I ask that you please provide attribution where possible. Or better yet, consider becoming a supporter or collaborator within this movement.
The reality is that I cannot continue to innovate and publish these ideas freely without support from the broader education community. When educators adopt these frameworks without acknowledgment, it becomes difficult to sustain the research and development that creates new insights for all of us.
This is about building a sustainable ecosystem where innovation in AI literacy can continue to flourish and reach the educators and students who need it most.
The Deeper Benefits
The sparring partner approach does more than just improve AI interactions—it develops transferable skills that students will need throughout their careers.
Critical Evaluation: Students learn to assess information quality in real-time, a skill that applies far beyond AI interactions.
Clear Communication: The need to communicate effectively with AI translates to better communication with humans.
Intellectual Confidence: Students develop the ability to engage with sophisticated tools while maintaining their own agency and judgment.
Metacognitive Awareness: The reflection required for effective AI sparring builds students' ability to think about their own thinking.
The Choice Is Ours
Every day, students are developing their relationship with AI. They're forming habits, building mental models, and establishing patterns that will shape how they engage with AI throughout their lives.
We can let them default to the butler relationship where AI does work for them and they gradually become more dependent on it.
Or we can help them discover the sparring partner relationship, one where AI becomes a catalyst for demonstrating and developing their own capabilities.
The choice isn't whether students will use AI. They're already using it, and they'll continue to use it more. The choice is whether we'll help them use it in ways that make them stronger thinkers, better communicators, and more confident learners.
The sparring partner approach doesn't eliminate the risks of AI in education—but it transforms those risks into opportunities for deeper learning. And in a world where AI capabilities continue to evolve rapidly, helping students develop a healthy, productive relationship with AI might be one of the most important things we can do for their future success.
Join the Movement
Ready to be part of the Critical AI Literacy movement? Fill out this quick form to let us know how you’d like to get involved. Here’s what you can do:
Implement & Share: Try the sparring partner approach in your classroom and share your results with the community. I have highlighted a number of educators who have tried this concept in their classrooms here on the blog — both independently and in partnership — and would love to highlight your experiences as well.
Amplify the Message: When you present or publish work inspired by these ideas, include attribution to help others find the source.
Contribute Your Insights: Share your own discoveries and adaptations—this work gets stronger when more voices join the conversation.
Become a Community Supporter: Whether through formal collaboration, financial support, or simply spreading the word, help ensure these resources remain freely available.
This movement succeeds when educators support educators. If these frameworks and experiences have changed how you think about AI in education, help ensure it can continue evolving and reaching the teachers and students who need it most.
The sparring partner metaphor is part of a broader Critical AI Literacy framework that helps students develop the skills needed to collaborate effectively with AI while maintaining their independence as thinkers and learners.