The Kids Are All Right: What Students Actually Say About AI

One student in a Jisc survey described a personal rule: anything that reached her Word document had to be written in her own sentences.

It is a small, clear boundary. It also offers a different starting point for discussing students and AI than another headline about cheating.

In this episode of Adjunct Intelligence, Dale Leszczynski and Nick McIntosh examine what students say about their own AI use, learning and future work. The conversation deliberately turns towards the findings that often receive less attention: students' doubts, self-imposed limits and desire to retain ownership of their thinking.

Using AI while worrying about it

The studies discussed show widespread adoption alongside considerable caution. Students describe limiting their use because they value learning, want to complete work themselves and worry about becoming dependent on a tool they know can be unreliable.

Dale shares a similar experience from a student advisory group. Students were already discussing cognitive offloading and the need to develop enough subject knowledge to recognise when an AI system was leading them astray. Their concerns were specific, thoughtful and closely connected to what they hoped to gain from university.

The episode also confronts the tension between values and behaviour. Survey responses describe what people report about themselves. A student who values independent thinking can still take a shortcut under deadline pressure. The hosts acknowledge the increase in students reporting direct inclusion of AI-generated text in assessed work.

That keeps the argument grounded. Student agency deserves attention, and institutions still need to address behaviour that undermines learning or breaches assessment requirements. Both belong in the conversation.

Another tension concerns careers. Students are told that AI skills matter for employment while also hearing that the technology may reduce entry-level opportunities. They need support to navigate that uncertainty, alongside practical guidance about their immediate coursework.

Clearer guidance, with students in the room

Ambiguous policy leaves students to fill the gaps themselves. The episode describes guilt, uncertainty and even students changing their behaviour because they believe their university uses detection systems it does not actually have.

That atmosphere can consume attention that should be available for learning. Guidance becomes more useful when it explains the role AI can play in a particular task, the thinking students need to retain and the reasons behind those expectations.

One of the strongest examples comes from a student-led workshop in Vietnam. Students guided staff through the consequences of vague AI rules for people with different learning and access needs. A seemingly straightforward restriction looked different when a student relied on assistive technology to participate.

The conversation also returns to personal voice. Students notice when fluent AI-generated writing starts replacing the way they express themselves. Some are already creating their own safeguards, including forming a position before consulting a tool.

The episode's central invitation is to take that reasoning seriously. Universities can build clearer, more inclusive approaches by working with students who are already thinking carefully about these questions. Listening to them reveals a much richer picture than adoption percentages alone.

Related links

• Jisc: Student perceptions of AI 2025
• HEPI: Student Generative AI Survey 2026
• AI in Higher Education research project

Find out more

Find out more about the student voices, research findings and disagreements behind this conversation. Listen to the full episode or watch Dale and Nick explore what students are asking of universities on YouTube.

Listen to the full episode · Watch the episode on YouTube