“Can I ask you something?” a student inquired after class, lingering a bit longer than usual.
“Of course,” I replied.
She hesitated. “If I can use AI to do this assignment, what does that actually mean about the assignment?”
I paused. “What do you think it means?”
“I think,” she said slowly, “it means maybe the assignment isn’t really measuring what I know. Or how I think.” We sat with that for a moment.
“So, if assessments are doing what they’re supposed to do,” I said, “what might AI show us?”
She smiled a little. “Probably the same thing it’s showing us now. It’s not creating the problem. It’s just making it obvious.”
That exchange has stayed with me, resurfacing in conversations with students, in faculty meetings, and whenever I pause to look closely at what an assignment is truly asking students to show us. In my work with students, colleagues, and schools, I keep returning to a hopeful idea: AI is not the enemy. Instead, it is a mirror, showing us where our assessments or ideas about learning may be falling short.
When a student can produce work that looks complete without explaining their choices, questions, revisions, or reasoning that shaped their work, AI has not necessarily created the problem. Instead, AI may simply be revealing that we have valued completion more than evidence of understanding.
Rethinking What Assessment Shows
Generative AI entered classrooms almost overnight, and students began experimenting with it before most schools had developed a shared language for how to approach it. In independent schools, where relationships and trust shape the learning environment, this shift felt especially complex. Our first instinct was understandably to protect academic integrity: What should students be allowed to use? How should they disclose it? How can teachers know whether the work truly represents a student’s own thinking?
These questions are important, and teachers and students both deserve clear guidance. At the same time, these questions do not take us all the way to the heart of the issue. Detection tools remain inconsistent, and an emphasis on monitoring can pull teachers into a role that feels out of sync with the trusting, curious classrooms we are trying to build.
More importantly, a focus on enforcement can draw our attention away from the questions that do matter most: Why are we using this assessment? What does this assessment allow a student to practice, and what does this assessment help us understand about that student’s learning?
I am hopeful that many schools are beginning to center those questions in thoughtful conversations about what counts as evidence of learning and how we can design assessments that invite students to make their thinking visible.
Prioritizing Process Over Product
One of the most promising shifts I’ve seen in my school, Miss Porter’s School (CT), is a renewed attention to process. Across disciplines, teachers are finding creative, practical ways to make the learning journey, and not only the final product, more visible.
In mathematics classrooms, this often means inviting students to make their reasoning visible. Drawing on Peter Liljedahl’s work in Thinking Classrooms, students explain not only what they did, but why they chose that approach. A correct answer is only one part of the story; the explanation allows us to see understanding, persistence, and growth in ways an answer alone cannot.
In science, some teachers are moving from predictable labs toward student-designed investigations in which students pose questions, justify their approach, and explain how the evidence shaped their conclusions. AI may support background research, but it cannot replace the decisions students make along the way.
In my language classrooms, I now rely even more on conversation. Using station rotation models inspired by Catlin Tucker’s (education expert, author, and international speaker) work, students might read a short Latin passage, collaborate to sequence or retell the story, and prepare a brief response while I meet with small groups. I ask students how they made sense of a passage, what evidence supported their interpretation, and where they still feel unsure. I love these moments because I can hear students constructing meaning in real time and because these moments remind us that language learning is fundamentally about communication.
In English, writing assignments are opening up in similarly meaningful ways. Reflective writing, multimodal projects, and revision across drafts create space for students to develop their own voices. A student might pair a literary analysis with a podcast, visual essay, or recorded commentary and then reflect on what each medium allowed them to express. The final piece still matters, but so does the learning path that brought the student there.
Giving Feedback and Teaching Integrity
Once we make thinking more visible, feedback becomes both more important and more human.
When students can use AI to generate ideas quickly, what they often need most is a trusted person who will help them shape, question, and deepen those ideas. Feedback from teachers and peers creates space for students to refine their thinking, reconsider their choices, and discover possibilities they may not have seen on their own.
I am excited to see more teachers intentionally building time for short conferences, peer conversation, and reflection into the rhythm of a lesson. These moments are not add-ons; they are often where students feel seen and where some of the most meaningful learning happens.
A tool can generate polished language, but it cannot notice the hesitation in a student’s voice, celebrate a breakthrough, or ask the follow-up question that helps an idea grow. It cannot replace the feedback that grows from a relationship or the ongoing conversation between a student and teacher that shapes real learning.
Relational work matters just as much when we talk about academic integrity. For many students, using AI does not feel like cutting corners; it feels like using another resource, not so different from Google or Wikipedia. Without guidance and opportunities to practice, we leave students to navigate choices that are far more complex than they may initially appear.
Our task, then, is not only to enforce integrity but also to teach it intentionally. We can help students connect honest decision-making to the thinking processes we are already asking them to make visible.
In a history classroom, for example, students might reflect on how they used sources, including AI, and describe the choices they made along the way. In English, students might annotate revisions and explain what prompted them to rethink or refine an idea. In mathematics, students might compare their own solution with one generated by AI, identifying where the approaches align, where they differ, and what each reveals.
These practices will not eliminate AI misuse entirely, but they do something more lasting: They bring student thinking into the open. These practices help young people develop judgment, reflect on their decisions, and take greater ownership of their learning, which are skills that will serve them far beyond a single assignment.
Recognizing a Moment of Possibility
AI will continue to evolve, and it will be increasingly present in students’ lives, both in school and in the paths they choose after graduation. We may not be able to control the pace of that change, but we can shape how students encounter it: with curiosity, care, and a clear sense of purpose.
The encouraging news is that not everything needs to change at once. A department can revisit a single assessment. A teacher can redesign one assignment. One honest conversation with students can lead to a clearer understanding of how they approach their work and what support they need.
These small shifts are especially possible in independent schools, where relationships and flexibility create room for thoughtful experimentation. We can revisit curriculum, learn alongside colleagues across departments, and keep students’ growth, not the technology itself, at the heart of the conversation.
AI has not changed what we value most in education. If anything, it has brought those values into sharper focus through the ease of production. Meaningful learning still grows from curiosity, reflection, connection, and the willingness to think through complexity.
If AI is holding up a mirror, I hope we are willing to look closely, not with fear, but with curiosity, and ask what the reflection can teach us. This is our opportunity to strengthen assessment, deepen relationships, and recenter the student learning and growth that brought so many of us to education in the first place.