Summary

• Data points to a possible anti-AI majority of students, at least for use in their educations
• There are big implications for student associations, who have to do more to understand nuances
• Considerations include changes in office use, different advocacy approaches, and more criticality in academic assessment conversatoins
• There is also a subtle need to use AI to shape the minds of students to vote for incumbents

New data

CASA-ACAE’s most recent statistical release, based on scientific polling has some new statistics on artificial intelligence in post-secondary. In particular, they’ve released data on student’s preferences for AI integration within their education, and the findings are that students are not about it.

First, in the last two years, AI adoption declared students has functionally plateaued at around the 40% mark, even when it’s in an area with no risk of academic misconduct, like summarizing readings. The data is contrasted against questions in a 2024 survey by CASA, and while in some areas AI use has increased, but with regard to using generative AI, it has gone done since 2024.

Secondly, asked to rank what concerns them for Canada’s future, AI functionally equals that of climate change.

Thirdly, only one in five students want deep AI integration into courses and learning tools. And this is the piece worth considering today.

The fact that 79% of students reject AI explicitly suggests that student associations should become anti-AI writ-large.

So what?

Just asking questions (would be a good start)

I think this survey question immediately makes me want to see other student associations or supraunions jump into getting more detail. As someone who had a hand in crafting this question, I do feel like I probably fell into the trap of giving people a median answer that was too agreeable and not specific enough. Asking further questions about what that middle category looks like and digging into more detail is probably going to elucidate if there is a functionally anti-AI majority in the student population. For example, maybe there are a lot of students who want AI training, but want it heavily segmented from the rest of their leaning. Or maybe they trust themselves to use it but not their professors. These are two very different beliefs that should be disentangled.

The second question is digging into the motivations behind students’ beliefs. Do they have a philosophical opposition to AI use? Is that opposition environmental or based on rights of creators? Or maybe their opposition is based on not wanting to see the AI majors (the companies) profit? Or maybe they see AI as detrimental to their own learning and quality of instruction. In any case, doing some digging will probably help shed some light on this, but should probably be done fast, snice the landscape is changing fast.

Finally, and I don’t blame CASA for this, as they have to keep the lengths of these documents reasonable so they can’t cross-tab everything, but they didn’t include any crosstabs here, and whether there are regional or college-university distinctions here may influence how different associations think.

Operations

If data reveals that students’ opposition to AI use is philosophical, then there is probably a question to be asked about the use of AI in the student associations themselves. Associations may want to review their internal use policies not for utility, but for philosophical alignment with the expectation of how their membership expects them to ask. This, as in many things, will pit pure representation against efficiency gains, and Boards of Directors may be called upon to make that judgment call, but they can only do that if associations have robust internal understandings of where things already sit.

Shaping Advocacy

If students are fundamentally against AI use, student associations might want to revisit whether they are advocating for good AI implementation, instead pivot to the conversation about whether AI should be implemented at all, or at least minimizing where it will be.

Given that CASA likely had this data a little longer, it did make me curious why they were praising an AI literacy initiative. Granted this policy was coming whether they like it or not, and there are arguably up to 53% of students who might like it, but it was already announced, and supporting government when it spends money of students is usually a good approach if the spending isn’t actively harmful. However, it probably makes it harder for them to make the pivot if they make the determination that they’re going to be generally anti-AI in the near future, especially in light of the new Mitacs funding for AI positions announced late last week.

Assessing academics

If anyone is choosing to do the hard but important work of assessing academics and being a partner in collegial academic governance, you probably want to integrate AI use into your assessments, and doing so as a critical-by-default approach.

Government advocacy

If you’re doing government advocacy, are anti-AI and crucially, have the O.K. from your organizations to do so, relay the data to politicians you’re meeting with! You probably won’t use it to shape recommendations too much in the near future, unless something big changes in student priorities, but it probably doesn’t hurt to have a statistic or two in the back of your head, because I bet as student associations hit their provincial and federal advocacy weeks this year, the politicians will be asking.

Politicians will be asking because they are arguably the most AI skeptical category of people out there, and for good reason; an assessment of AI job-exposure data shows that they are one of the first careers to be put totally out of work.