During our Customer Discovery Sessions in Oslo, we didn’t just talk with admins and educators — we also sat down with students to co-create the future Canvas mobile experience as well as AI experiences for learning.
At the University of Oslo, students from different faculties and year levels joined us for two focused sessions:
- A Student Dashboard co-creation workshop
- A Canvas AI Experiences session on how LLM-based tools should (and shouldn’t) show up in their learning
Their message was very clear: they want a dashboard that helps them feel in control of their academic life, and AI that acts as a learning companion.
What students really need from their mobile dashboard
We kicked off with a short interactive survey before moving into hands-on design activities. Students then worked in small groups to build their ideal mobile dashboard using printed, predesigned widgets and layouts.
The daily reality: logistics + deadlines
Students told us that their biggest mental load comes from:
- Keeping track of assignments and deadlines
- Figuring out where they need to be and when (changing rooms, schedules, lecturers)
- Preparing for lectures and exams across very different course structures
Most use the Canvas app and their phone’s calendar, alongside local tools like a “My Studies” app. Many also juggle multiple other apps for reading lists, transport, messaging, and more.
When we asked what they want to see first thing in the day, the answer was almost unanimous:
“What do I have today?”
“What is urgent or overdue?”
Customization is a must-have, not a nice-to-have
One medical student summed it up best:
“I have one big course and no grades – half of this dashboard just wouldn’t exist for me.”
Students want to choose which widgets matter, decide what’s expanded or collapsed by default, and reorder components to match their own priorities.
Simple, minimal, and “click-to-expand”
Across all groups, the preferred design philosophy was clear:
- A clean, minimal, modern interface
- Scannable summaries (for today / this week)
- Ability to tap to see more detail, rather than having everything expanded by default
This “click-to-expand” pattern was their preferred way to balance minimalism with the need for rich information. Several groups also designed swipeable rows for courses and assignments to reduce vertical scrolling.
Design with student well-being in mind
One of the widget showed motivational texts, another had comparison with course average. Students highlighted the emotional impact of what we choose to show by default:
- Comparative metrics like course averages can feel stressful and create competition.
- Motivational elements, like Canvas pandas and short “you can do it” messages, are polarising — some find them friendly, others see them as clutter or irrelevant.
They ask: anything that could increase stress (like grade comparisons or “gamified” elements) should be optional and opt-in, not part of the default experience.
How students want AI to show up in Canvas
In a separate session, we explored LLM-based AI Experiences with students from a mix of disciplines. The group already uses tools like ChatGPT and Gemini regularly — for summaries, drafting, language help, coding, and brainstorming.
But when it comes to AI in their courses, they were cautious.
AI as a learning tool, not just another exam
Students were interested in AI when it:
- Helps them learn faster (e.g. summarizing, clarifying, giving examples)
- Supports practice and iteration, not one-shot grading
- Gives structured progression — like moving from basic tasks to more complex conversations, similar to how language apps scaffold difficulty
They were much less enthusiastic about AI being used as a hidden layer in formal assessment, especially if teachers only see a small part of the interaction.
Trust, privacy, and guardrails
Even though they use AI a lot, students were clear about their boundaries:
- They don’t fully trust AI-generated study materials unless they or their teacher created or approved them
- They want to know where their data goes, whether conversations are stored, and who can see them
- They value concise, clear outputs, not long, over-explained responses
What this means for the Canvas student experience
Taken together, these student-led insights give us a clear direction for the future of the Canvas mobile dashboard and AI Experiences.
For the Student Dashboard
We’re using this feedback to guide ongoing and future work, including:
- Personalization by design
- A widget-based dashboard where students can add, remove, and rearrange components
- Widgets that remember their expanded/collapsed state
- Clarity-first layout
- A “morning overview” focused on today / this week, lectures and deadlines
- Compact, modern UI with “click-to-expand” rather than information overload
- Well-being-aware defaults
- Removing unnecessary stressors from the default view
- Making grade-related comparisons and motivational elements opt-in
- Set the right priority
- Based on the selected widgets the first increment of the dashboard is finalized: This Week view, Course-Groups view, Hello-widget, Calendar - Today widget and of course the dashboard customization.
For AI Experiences
Student and admin feedback in Oslo is also shaping how we think about AI in Canvas:
- AI positioned as a learning companion, not just an exam tool;
- Emphasis on progression, reflection, and evidence of growth, not only final answers;
- Strong focus on privacy, transparency, and guardrails, in line with regional expectations.
Thank you, Oslo 
A huge thank-you to the students who joined us, shared their honest experiences, and literally rearranged our dashboard ideas on the table. Students designs, quotes, and strong opinions are directly informing how we evolve the Canvas mobile experience and our AI roadmap.
Finally, a special thank you to the University of Oslo for hosting our sessions and for supporting this critical co-creation effort. We greatly appreciate the partnership and the opportunity to engage with your diverse and thoughtful student body. The insights gathered within your walls are invaluable as we design the next generation of learning experiences.
We’ll keep sharing how these insights turn into product changes — and we look forward to coming back to validate the next iterations.
