Learning with AIEducation qualityResearch
Adaptive feedback as the key to responsible GenAI use in higher education
This week saw the publication of a relevant study by Kangwa et al. (2025), based on 137 studies on the use of generative AI (GenAI) in higher education worldwide. One important theme emerges: how do we make sure students do not become dependent on GenAI, but instead use it to learn better? In this blog I focus on one powerful tool: adaptive feedback.
🚨 The risks of GenAI dependence
When students use GenAI without guidance or structure, risks can arise: 🔺 Reduced development of academic skills
Critical thinking, creativity and problem-solving ability can decline when students rely solely on AI output instead of analysing things themselves.
- Lower motivation and learning outcomes
Kangwa et al. observe that less active learning goes hand in hand with reduced motivation and, ultimately, poorer results.
- Passive and superficial learning strategies
Without reflection or deeper engagement, learning stays on the surface. Copying AI texts word for word prevents deeper understanding.
✅ The power of adaptive and personalised feedback
Yet that same GenAI also holds the solution. Kangwa et al. argue for an active role for GenAI in the learning process, in which the model does not just answer, but also guides.
Adaptive feedback:
- 🧠 Provides insight into strengths and weaknesses
- 🌱 Encourages self-reflection, engagement and motivation
- 🛠️ Helps students use more deliberate learning strategies, such as planning, monitoring and adjusting
In short: feedback that adapts to the individual student promotes self-regulation and prevents dependence.
🧭 How do you encourage GenAI feedback that really works?
Would you like to get started with adaptive GenAI feedback as a lecturer or programme? Then these are important principles:
👉🏽 Upload rubrics, learning objectives and clear criteria
When GenAI knows what matters, it can give feedback that fits the assignment, for example on structure, argumentation or use of sources.
👉🏽 Teach students to apply ‘prompt design’
The more specific the question to GenAI, the more valuable the feedback. Let students experiment with prompts focused on process and reflection.
👉🏽 Plan deliberate moments of reflection
Make sure students do not simply adopt feedback without thinking. Discuss what is valuable, what is not, and how they can use follow-up prompts to go deeper.
🎓 Recommendations for education
Would you like to work responsibly with GenAI in the classroom as a programme or team? Then consider these guidelines:
- Prevent passivity – Provide active guidance and ask critical questions.
- Ensure process-oriented feedback – Ask: “How did you arrive at this answer?”
- Offer training and support – Explain how GenAI works, technically and ethically too.
- Monitor and evaluate – Keep track of how students use GenAI and adjust where needed.
- Emphasise cultural relevance – Make GenAI fit the language, context and background of students.
- Support critical thinking – Have students assess, improve or add to AI output. That way the human stays in charge of the learning process.
- Want to know more?
Read the full article by Kangwa et al. (2025):
Would you like to use adaptive AI in your teaching practice? Or strengthen feedback processes with local AI? Then take a look at my feedback literacy training or the AI workshops at Education Oracle.
