How to reclaim humanity in the AI classroom

Nine in 10 UK undergraduates who took part in a recent Higher Education Policy Institute (Hepi) survey said they had used GenAI in their assessments. Students can now produce polished assignments without engaging deeply with the underlying concepts. The result is a growing risk that automation will hollow out the process of learning itself, turning intellectual struggle into a performance of intelligence. Universities must therefore reaffirm a critical truth: we cannot automate or outsource learning. Writing, researching and reflecting are all key to understanding.

Rediscovering human fundamentals

To navigate the opportunities and risks of GenAI, we must ground innovation in the principles of human learning. One starting point is to use AI to support, rather than replace, cognitive development. Well-designed AI tools can reduce cognitive overload by providing examples, structured explanations or formative feedback, allowing students to concentrate on reasoning and interpretation. Yet this potential is only realised if students are explicitly taught metacognition, the ability to think about their own thinking.

Metacognitive learners become discerning technology users because they are trained to notice how AI shapes their thinking. We can develop this through simple but powerful classroom practices. For example, ask students to keep brief AI reflection logs in which they record what they used a tool for, what it returned and how they revised or challenged the output. Another effective exercise is “compare and critique”, where students complete a task both with and without AI and reflect on differences in reasoning, depth and confidence. You can also use “think-aloud” activities, modelling how to evaluate an AI response in real time by questioning its assumptions, evidence and tone. These routines turn AI from a shortcut into a scaffold and help students build awareness of their own learning habits. In doing so, institutions protect the reflective capacity that lies at the heart of intellectual development.

• Spotlight guide: Bringing GenAI into the university classroom
• Three ways to use ChatGPT to enhance students’ critical thinking in the classroom
• Is critical thinking the answer to generative AI?

Clarity and connection in teaching

Amid digital change, the foundations of effective teaching remain remarkably consistent. Students need clarity: about not just what they are learning, but why it matters. Make learning goals transparent, and assessments a continuation of learning rather than a final judgement. When feedback, whether human or AI-generated, helps learners improve, it transforms assessment from a grading exercise into an act of growth.

Learning also remains profoundly social. In an era marked by personalised digital feeds and algorithmic recommendations, universities must remain spaces for dialogue, disagreement and shared discovery. Group discussions, collaborative projects and peer review cultivate empathy, intellectual curiosity and resilience. Even the most advanced machine cannot yet replicate these qualities. They are the hallmarks of meaningful education.

AI as a cognitive co-pilot

The most effective use of AI in higher education will come from what might be called extended cognitive hygiene, a disciplined approach to selecting and using digital tools. Every introduction of AI should serve a clear pedagogical purpose: to prompt reflection, stimulate creativity or alleviate cognitive load.

AI literacy is central to this process, and it must be taught explicitly rather than assumed. Educators can begin by building short “verification tasks” into lessons, requiring students to check AI-generated facts against trusted sources and to identify inaccuracies or gaps. Students can explore bias through comparison activities, where learners analyse how the same prompt produces different responses across tools. Asking students to label claims as “evidence-based”, “speculative” or “unsupported” also builds habits of critical evaluation. Over time, these structured exercises turn scepticism into skill, enabling students to engage with AI confidently while retaining academic judgement and integrity. 

Likewise, we should view prompt design as a new way to develop critical thinking. Prompt-redrafting, where students submit an initial question, receive feedback and then refine it for clarity, depth and purpose, is an effective approach. Another is the “question ladder”, where learners progressively improve a vague prompt by adding context, constraints and criteria for success. 

Thinking aloud while constructing a high-quality question helps students understand how intention shapes outcomes. Over time, these practices develop curiosity and precision, showing that good questions are not accidents but products of disciplined thought. AI should function as a cognitive co-pilot, extending rather than substituting human capability. This approach enables universities to combine digital fluency with intellectual integrity.

The future of human-centred learning 

GenAI does not signal the end of learning; it offers an invitation to rediscover what it truly means to learn. The task is to blend the precision of machines with the empathy, creativity and ethical awareness that only humans can provide.

With GenAI constantly improving its capabilities, the future of higher education depends on cultivating intelligent, reflective and compassionate humans. The challenge ahead is not to outsmart AI, but to outgrow it.

Patrice Seuwou is an associate professor of learning and teaching and director of the Centre for the Advancement of Racial Equality at the University of Northampton.

If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.