With the rise of generative AI tools like ChatGPT, the conversation around academic integrity in schools has reached new urgency. For Noémie Verhoef, a philosophy teacher at Cégep de Victoriaville, simply banning technology or reverting to paper-based assessments won’t solve the challenges AI brings. Instead, she believes the solution lies in rethinking how we teach and evaluate students.
During a recent webinar hosted by the Association québécoise de pédagogie collégiale (AQPC), Verhoef shared how her approach has evolved—from focusing on detection to prioritizing thoughtful guidance. Her transformation began in winter 2024, after encountering a spike in plagiarism in a long-standing midterm assignment.
“ChatGPT murdered my evaluation,” she said candidly.
Initially tempted to return to traditional formats, she instead examined the learning context. “Plagiarism existed before AI,” Verhoef explained. “It’s not just about banning tools—it’s about teaching students how to use them responsibly.”
Verhoef redesigned her assignment—a half-essay on a philosophical question—over three sessions. Her new approach allows limited, supervised AI use while guiding students through authentic research and writing processes. Key changes included:
Students now complete their essays in a teacher-provided Teams document with tracked changes, ensuring transparency and accountability.
The outcomes speak volumes: plagiarism rates dropped from 10 cases among 89 students to 4 among 118. Even non-submissions fell from 10% to just 4%. More importantly, classroom dynamics improved. Students are now more engaged, with richer discussions and a deeper investment in their learning.
“Instead of playing detective, I went back to being a teacher,” said Verhoef. “Pedagogical problems deserve pedagogical solutions.”
She’s skeptical of purely technological fixes like AI detectors, which she sees as part of a never-ending race. Instead, she champions careful planning, meaningful activities, and a relationship-based approach.
“When you have a good relationship with your teacher, it’s harder to cheat.”
Verhoef’s position on AI is nuanced. While she sees potential—such as AI acting as a tutor or summarizer—she’s clear that technology shouldn’t replace students’ intellectual work.
“AI is fine, but humans are better,” she quipped.
Her strategy fosters student autonomy and emphasizes responsibility as a path to freedom. This includes group work, collaborative knowledge-building, and reflective practice.
“Teaching freedom means teaching responsibility,” she noted. “Students must be given clear boundaries so they can reclaim their space. They are at the centre of their learning.”
In fall 2025, Verhoef will make her learning journal a graded component to encourage more thoughtful engagement. Students will also shift away from using AI to find arguments, opting instead for in-class peer discussions. AI will still be permitted for focused tasks like summarizing research.
“Before AI, students could formulate philosophical questions,” she said. “There’s no reason they can’t still do it today.”
Verhoef’s approach has not only improved outcomes but also restored purpose in her teaching—while respecting her students’ capabilities and independence.
📺 Watch the AQPC webinar with Noémie Verhoef (in French): https://www.youtube.com/watch?v=bQrfXMZ3M4E
If you are subscribed, log in to access exclusive member content!