Jinso logo
← Back to Briefings
July 22, 2026

Field Notes

Half of Latin American Students Want AI Feedback. One in Five Faculty Give It.

LATAM6 min read·By: Jinso Labs

The Digital Education Council's AI in Higher Education LATAM Survey 2026 turned up a specific, narrow gap worth sitting with longer than the adoption headlines it came attached to. Half of students across the 29 surveyed institutions support instructors using AI to give feedback on their assignments. Only 19% of faculty currently do it. That's not a small gap in a large survey. It's a 31-point disconnect between what students want and what's actually happening in the classroom, and it's worth asking precisely why it exists before assuming the obvious answer.

The obvious answer — faculty resistance — doesn't hold up against the rest of the same survey. Only 12% of faculty see AI as a threat to their role. 94% expect to be using it more in their teaching going forward. Those aren't the numbers of a faculty body digging in against a tool they distrust. Something else is keeping the 50% and the 19% apart.

What the Numbers Actually Show

  • 50% of students support AI-assisted feedback on their work; only 19% of faculty currently provide it
  • 12% of faculty view AI as a threat to their role — low enough to rule out resistance as the primary explanation
  • 94% of faculty expect to use AI in future teaching, consistent across years of experience
  • 79% of faculty are already actively engaging with AI in some capacity, just not yet in feedback delivery specifically

Hector G. Ceballos, Head of the Living Lab and Data Hub at Tecnológico de Monterrey's Institute for the Future of Education, made a similar point about why this gap deserves institutional attention: how teachers and students actually experience AI in feedback and assessment is exactly the evidence institutions need to shape policy and guidelines going forward, not a detail to note in passing.

The Real Obstacle Is a Skill, Not an Attitude

A complementary piece of DEC research gets closer to the mechanism. The Digital Education Council's AI Skills Opportunity Map, an initiative developed with Google.org, analyzed AI's effect on eleven job families and found that education and teaching professionals show the highest AI embeddedness of any profession studied — and identified assessment design specifically as one of five skills that AI has structurally redefined for educators. Giving useful feedback on AI-assisted work isn't the same task as giving feedback on work produced without AI involved. A faculty member now has to design an evaluation that can tell the difference between demonstrated understanding and well-formatted AI output, and that's a genuinely new skill, not a faster version of an old one.

That reframes the 19% figure. It isn't 19% of faculty choosing to give feedback while the rest decline. It's closer to 19% of faculty having already rebuilt their assessment practice for an AI-assisted classroom, while the other 81% are still teaching with tools built for a classroom that doesn't exist anymore. The DEC research is explicit that this skill needs to be built in sequence — AI introduced after domain expertise is established, not as a replacement step — which means the fix isn't a single training session. It's a structured, ongoing capability faculty need time and support to build.

Where the Structure Has to Come From

Institutions waiting for this gap to close on its own are waiting on individual faculty members to independently reinvent assessment design, one classroom at a time, with no shared standard for what "AI-assisted feedback" should even look like. That's a slow and inconsistent way to close a 31-point gap, and it puts the entire burden on the person least equipped to solve it alone.

This is precisely the layer Jinso's teacher training component and AI Agent are built to sit underneath. Unlimited live and on-demand training gives faculty the specific, rebuilt skill the DEC research says they're missing — not generic AI literacy, but structured practice in designing assessments that can actually evaluate AI-assisted work — while automated assessments already AI-graded and posted to the institution's LMS give faculty a working model of what that evaluation looks like before they have to build it from scratch themselves. An institution that hands faculty a working example of AI-assisted assessment, rather than a policy memo telling them to figure it out, is closing a skill gap instead of just naming one.

The 31-point gap between what Latin American students want and what their faculty are currently delivering isn't a motivation problem waiting on more encouragement. It's a skill gap waiting on the training and infrastructure to close it — and the institutions that build that infrastructure first won't need to convince their faculty of anything. The demand is already sitting in the survey data, unmet.

References

← Back to all briefings

More Briefings