Beyond policing AI: How universities can teach better in the age of GenAI

Generative AI has made life difficult for universities. Tools like ChatGPT can now write essays, summarise research, create images, and produce polished work in seconds. This raises a big question: how can teachers know what students have actually learned?

This article argues that universities should stop focusing only on catching AI use. Detection tools are often unreliable, can wrongly accuse students, and may be unfair to students who write in English as an additional language. Since AI keeps improving, policing it will always be a losing race. Rather than focusing on blocking or detecting AI, teachers can design learning experiences that use AI as one part of a wider process of inquiry, reflection, and exploration.

Working with the idea of a rhizome (a root system that spreads in many directions) is a useful way to think about learning. Learning does not happen in a straight path. Students make connections, follow unexpected ideas, hit dead ends, change direction, and build understanding through experience. AI is very good at producing common, polished answers. But it does not have lived experience. It cannot walk through a neighbourhood, reflect on a personal memory, take part in a difficult group discussion, or explain how its thinking changed over time. That is where better pedagogies and assessments can begin.

Instead of only grading final essays or projects, teachers can ask students to show their learning process. For example, students might submit their sketches, AI prompts they used, notes on what AI got right or wrong, drafts showing how their ideas changed, reflections on dead ends or mistakes, and evidence from real-world observation or conversation.

Teachers can also design tasks that require students to adapt. A project might suddenly shift audience, format, or viewpoint, forcing students to rethink their work. Students could also begin with an AI-generated answer and then ask: What did AI miss? In this model, the teacher is not just enforcing the rules but guiding students to question, connect, reflect, and explore.

The key message is for universities and teachers is towards paying more attention to how students move through a learning process: how they use tools, respond to uncertainty, make connections, and revise their thinking. This also depends on nurturing conditions for communities of learning, where students and teachers can share questions, test ideas, reflect together, and support different paths of inquiry.

The better question is not “Did the student use AI?”

It is: “How did the student participate in a wider learning process that included AI, peers, teachers, and other sources of knowledge?”

To read more here. 

Tsao, J. (2026). From mitigation to exploration: Reimagining teaching and learning with generative AI in higher education. Innovations in Education and Teaching International, 1–15. https://doi.org/10.1080/14703297.2026.2726895