CEGEP vs University AI Policy – Same Problems, Different Contexts
After presenting at three different PD days this spring, I’ve been comparing CEGEP and university approaches to AI academic integrity. The differences are more interesting than I expected.
CEGEP challenges specific to Quebec: students are 17-19, transition-age between high school rule-following and university self-direction. AI policy is landing during this maturity transition, which complicates enforcement. CEGEP programs are 2-3 years, so policy coherence across cohorts matters more than at a 4-year university.
University challenges: graduate students expect more autonomy, which creates tension with disclosure requirements. Research contexts blur the line between AI as tool vs AI as collaborator even further. Turnitin’s AI detection is more likely to be wrong on specialized academic writing.
What’s consistent: neither system has a workable standard of evidence for AI-based academic dishonesty findings. probabilistic detection scores are not evidence by any formal standard.
The institutions getting this right in both contexts are moving toward transparency-and-demonstrate rather than detect-and-punish.
5 Replies
Join the discussion.
Log In to Reply
as a student who went through CEGEP last year: the policy confusion is real. different teachers in the same program had completely different rules and none of them matched what was in the official program guide. what you learn is to ask before every single assignment, not to internalize any principle.
the CEGEP-specific transition age point is one I haven't seen raised in the literature. sharing this with colleagues. the fact that students are navigating AI policy at the exact age where they're also transitioning from rule-following to self-directed learning is a genuine developmental complication.
to Greg's point: the CEGEP context has an additional layer because many students are transitioning from a secondary system that had one AI policy to a post-secondary environment that may have a completely different one - and then transferring again to university. three policy environments in four years. the inconsistency isn't just institutional, it's longitudinal.
the administrative challenge is that we can establish institution-wide principles but we can't prescribe pedagogy. which means the gap between policy and practice will always exist. we're working on a shared framework this fall that at least gives teachers a common vocabulary.
Après vingt ans en CEGEP, la distinction que tu fais sur la maturité est exacte. Les étudiants au cégep cherchent encore des règles claires tout en commençant à tester leur autonomie. Une politique d'IA qui dit simplement "divulguez et démontrez" demande un niveau d'autorégulation que beaucoup n'ont pas encore développé. Le CEGEP a besoin de plus de soutien structuré que ce que les universités offrent.