FEDERATION FOR THE HUMANITIES AND SOCIAL SCIENCES
Three Quarters of Cégep Teachers Have Changed the Way They Evaluate Students Due to Rise in AI, Says Montreal Researcher

With research showing a rise in post-secondary student dependence on artificial intelligence (AI) for assignments, increasing the risk of plagiarism, a new study indicates that 74% of Cégep teachers across Quebec have changed their method of evaluating students' performance.

In fact, almost a quarter of those teachers have come up with creative ways to tackle the issue, often dubbed as "AI-giarism" and in doing so are helping to boost students' cognitive thinking skills, including raising their awareness of the ways in which they use technology to support their education.

That's the message of Dr. Lisa Giachini, a teacher and researcher at Cégep Édouard-Montpetit, who will present findings of the study at the upcoming Big Thinking Summit: Inflection Point, a national academic gathering focused on the humanities and social sciences, taking place June 9 to 11 in Edmonton.

Organized by the Federation for the Humanities and Social Sciences, the Big Thinking Summit: Inflection Point brings together scholars, policymakers, community leaders, and institutional partners to examine the urgent questions shaping Canada's future. Featuring keynote conversations, research presentations, evidence-focused sessions, and networking events, it is designed to advance the role of humanities and social sciences knowledge in public life.

At the event, Dr. Giachini - lead researcher on the project - will share details of the province-wide study, which surveyed 345 teachers at 27 Cégep campuses via an online questionnaire. Twenty-seven of those teachers at 10 Cégeps were then interviewed to get a more in-depth understanding of their practices. Insights shared at the Big Thinking Summit will explore how teachers representing all school subjects conduct student evaluations and if their assessments have changed due to AI.

"With AI being so accessible to students today, schools must work hard to protect academic integrity against plagiarism, keep the evaluation process fair, and ensure students' cognitive development - an ability that AI can hinder," Dr. Giachini said.

Her study found that while 26% of the Cégep teachers queried haven't made any changes to their evaluation practices - believing their current evaluation methods to be adequate or the risk associated with plagiarism to be manageable - about half (49%) of respondents have adopted more strictly controlled methods to assess students. This includes administering in-class tests only and enforcing stricter exam guidelines and penalties for plagiarism.

The remaining 25% of teachers have completely transformed their student assessments, finding creative ways to make their evaluations more complex or requiring students to use critical thinking or apply their knowledge to real-life situations, Dr. Giachini explained.

"This result was particularly surprising and highlights a striking phenomenon: in the face of the challenges posed by the AI, human imagination remains a powerful driver of adaptation," she said. "It's easy to maintain the status quo or to test in a controlled environment with no access to technology, but this quarter of teachers puts the focus on measuring students' competency levels - not just their ability to perform tasks - which is the goal of education today."

As an example, she said many of these teachers evaluate their students' thinking processes by asking them to reflect on how they use AI to produce their work. "Having students respond to questions such as What prompts did you use? Why did you ask AI the questions you did? and What did you expect AI to give you based on your prompts? gives us the possibility to assess students' ability to plan, justify and evaluate their own actions - the real essence of competent performance in any domain," she said.

Dr. Giachini warns of the growing problem of cognitive offloading or shifting human thinking tasks to a machine. "Just as we teach students the basics of math before allowing them to use calculators, there are mental tasks that people need to learn before delegating them to AI. Once we have the foundation and move on to more complex goals, it becomes appropriate to use tools that handle the basics for us," she said.

"The findings of this study invite a collective reflection on the future of academic assessment," Dr. Giachini summed up. "AI in a way has forced us to be more human and work on our creativity, which is what differentiates us from AI, and I was astonished to see the high level of creativity of teachers in reconfiguring evaluations in a fairly short time period."

The Big Thinking Summit: Inflection Point is sponsored by the Social Sciences and Humanities Research Council of Canada, University of Alberta, University Affairs, Alberta Post Secondary Network, Canada Foundation for Innovation, Sage Journals, Pierre Elliott Trudeau Foundation, Universities Canada and Athabasca University Faculty of Humanities and Social Sciences.

Registration is open to the public, with one-day and three-day passes available. Visit www.federationhss.ca/big-thinking-summit-2026 to register and access the program of events.