How Are Students Using AI in Collaborative Learning?

Most course AI policies overlook how students actually work together. They focus on whether AI is allowed and how much students can use it. Yet AI can change how students develop ideas and resolve disagreements in group work. Without clear guidance, students are left to set their own rules with little means of enforcing them.

These findings come from Designing AI for Collaborative Learning, a new report by the Digital Education Council (DEC) in partnership with Pearson.

DEC conducted four online focus groups with 23 students from 21 higher education institutions worldwide. Participants ranged from undergraduates to doctoral candidates across more than 10 disciplines.

Collaborative learning depends on students building understanding together through discussion and debate. AI is now part of how students approach these tasks, with DEC’s AI in Higher Education Global Survey 2026 revealing that 88% of students use AI in their studies.

AI is reducing the debate that makes group work collaborative

Students described seven ways AI is changing group work. Many weaken the exchange of ideas that collaborative learning depends on.

Groups often arrive at meetings with similar ideas because students have prepared using the same AI tools. One participant observed that students come “with the same ideas because they are not our ideas”. With fewer differences to explore, groups lose opportunities to challenge each other’s thinking.

Some groups also ask AI to settle disagreements. Students may reach an answer faster, but they miss the reasoning involved in defending their own positions.

Other groups divide assignments into individual tasks completed with AI. One student then combines the outputs. Peers have fewer reasons to explain their work to one another. Polished submissions can also hide unequal effort, leaving engaged students to check and redo the work.

Students also reported growing pressure to use AI. They felt that instructors increasingly expect the quality and speed AI makes possible. 

One participant explained: “It’s not that we want to use AI, we can’t not use it”.

AI’s impact depends on when and how students use it

The same AI capability can support or weaken collaborative learning depending on its role in a task.

Students who use AI to generate their first ideas risk arriving at similar conclusions. When AI challenges ideas they have already developed, it can encourage deeper reasoning.

Students made clear distinctions across the stages of group work.

Before meeting, students welcomed AI for organising readings or scheduling discussions. They were more cautious about using it to generate initial ideas.

During group work, students valued AI that questioned their reasoning. They resisted AI that settled disagreements or took over discussions.

After completing a task, students welcomed AI feedback against assessment criteria. They preferred to address any gaps themselves rather than have AI correct the work.

The learning goal also changes what students consider appropriate. Slide design may be routine in medicine but a core skill in communications. Students in fields such as social work were particularly cautious about AI replacing human interaction. Business students more often saw AI as part of their future professional practice.

DEC’s analysis suggests that institutions need to define AI’s role around what each learning activity is intended to develop.

Students cannot enforce AI rules on their own

Students already try to manage AI use within their groups. Some agree to develop their own ideas before consulting AI. Others keep disagreements between group members and use AI feedback to test their conclusions.

These agreements depend on everyone choosing to follow them. Students doubted that peers would respect limits on AI use without clear course requirements or assessment. The pressure to complete work also makes informal rules difficult to sustain. When students are tired or facing deadlines, AI offers an easy shortcut.

Students reported better experiences when instructors designed AI use into the learning activity. They valued exercises where AI provided clues without revealing the answer. They also welcomed course chatbots that instructors monitored for accuracy.

Institutions need to define AI’s role in group tasks

The report identifies 15 emerging methodologies for integrating AI into collaborative learning and proposes six design principles.

The student findings also point to four decisions for institutions:

1. Define AI’s role at each stage. Specify when students may use AI during group work and what it may do.
2. Protect the learning goal. Identify which parts of a task require students to develop their own reasoning and contributions.
3. Use assessment to reinforce expectations. Require students to demonstrate their understanding through presentations or questions after submission.
4. Seek consent before using AI on student work. Students raised concerns about instructors uploading their work for AI feedback without permission.

General AI policies leave too much of collaborative learning to students to manage themselves. Institutions need course-level guidance that reflects how AI changes the learning process.

Read the full report, Designing AI for Collaborative Learning.

DEC members can also access Executive Briefing #033 in the DEC Member Area.