AI in Higher Education: Students and AI Project

What we are finding

Our project has produced papers, briefings, practical resources, commentary and presentations. This page brings that work together around the questions and insights emerging from the research.

Each section presents a key finding, explains why it matters for higher education, and links to the evidence and project outputs that sit behind it.

Looking for a particular paper, resource or presentation? Browse all project outputs here.

Evidence used across the project

  • a 2024 survey of more than 8,000 students across four Australian universities
  • a 2026 survey of more than 10,000 students across the same four universities
  • 20 focus groups involving 79 students
  • co-design labs involving 92 students, staff and industry participants
  • conceptual and methodological work that develops ideas and frameworks for educational practice

Emerging insights from the 2026 survey

In the first half of 2026, we repeated and extended our national survey to examine what has changed since 2024. More than 10,000 students shared their experiences and perspectives.

View the 2026 emerging insights snapshot

Broad findings

Click on a link to find out more:

1. Students are using AI in varied and selective ways

Students’ use of AI is more varied, cautious and purposeful than common public accounts suggest.

Students do not all use AI in the same way or for the same reasons. They make choices about when AI is useful, when its limitations matter, and when they would rather complete work themselves. Effective institutional responses therefore need to move beyond simple categories such as users and non-users, or responsible and irresponsible use.

Explore the evidence

Journal papers

Conference papers

Commentaries and media

Resources

2. Students actively negotiate integrity and ethical responsibility

Students actively reason about integrity and ethical AI use, sometimes holding themselves to standards that are more demanding than those formally articulated by their institutions.

Student decisions about AI are not explained adequately by whether a rule permits or prohibits a particular use. Their judgements also involve fairness, authorship, responsibility, learning and personal values. Guidance and assessment design need to support this judgement rather than assume that clearer rules alone will resolve uncertainty.

Explore the evidence

Journal papers

Conference presentations

Commentaries and media

3. AI is an emotional and institutional experience

Students’ experiences of AI involve uncertainty, anxiety, confidence, frustration and excitement. They do not simply ‘love AI’.

These emotions are not only responses to the technology itself. They are shaped by assessment conditions, institutional messaging, peer practices and uncertainty about acceptable use. How universities communicate about AI can therefore influence students’ confidence, sense of belonging and willingness to seek help.

Explore the evidence

Journal papers

Commentaries and media

4. Students value AI feedback, but teacher feedback remains distinctive

Students value AI feedback for its speed and availability, but generally place greater trust in feedback from teachers.

The key question may not be whether AI should replace teacher feedback, but how different forms of feedback can work together. AI may support immediate, iterative or low-stakes feedback, while teachers contribute contextual knowledge, disciplinary judgement and a valued sense of human attention.

Explore the evidence

Journal papers

Conference papers

Resources

5. AI use cannot be separated from pedagogy

AI use is entangled with how students learn, how teachers design activities, and how institutions organise educational practice.

Responding to AI is not only a matter of controlling tools or updating assessment rules. It also requires attention to the relationships among students, teachers, technologies, tasks, knowledge and institutional conditions. This shifts the question from “How should we manage AI?” to “What kinds of learning do we want to make possible?”

Explore the evidence

Journal papers

Presentations

Commentaries and media

6. Student partnership can shape more usable institutional responses

Students, staff and industry participants can work together to create AI resources that respond to the complexity of real educational settings.

Co-design moves students beyond the role of research participants or recipients of institutional guidance. It allows different groups to examine areas of disagreement, identify practical needs and develop resources that institutions can adapt to their own contexts.

Explore the evidence

Presentations