AI literacy cannot be bolted on to first-year courses

Policy statements, academic integrity warnings, student modules and permitted-use categories can leave students with the impression that AI literacy is mainly about compliance. They learn what is allowed, what must be acknowledged and what might get them into trouble.

These responses to generative AI, while necessary, are not enough. AI literacy is also the ability to decide when these tools might be useful, when they might be misleading, what needs to be checked and what responsibility the learner retains. In professional courses, preparing students to use AI well is even more significant. Future teachers, nurses, engineers, lawyers and business leaders will all make decisions in workplaces where AI is part of the environment. 

Start with learning that is already happening

In redesigning two first-year teacher education courses to incorporate AI literacy, I began with their purposes – not with where I could add a module. One course introduces students to the teaching profession and theories of learning. The other explores how human development affects teaching across childhood and adolescence.

• ‘Small changes in assessment design can make thinking visible’
• Your students use GenAI. Let’s redesign learning so thinking still happens
• The imperative for productive struggle

Both courses took a relational approach to teaching. I wanted students to experience belonging, dialogue, trust and clarity from the beginning of their degree. AI literacy had to sit within that architecture, not outside it. So, a more useful design question became: how can students learn to make responsible decisions about AI while they are learning the discipline?

This design shift changed the nature of both tasks and assessment. Instead of asking: “Did the student write this?”, the question became: “How was the student thinking?”

Use readings as a place to teach AI literacy

Course readings are one of the clearest places to do this work. Students are given guidance on how AI might support their engagement with complex academic material, but not replace it. For instance, AI can help students orient themselves to a dense reading, clarify terminology, summarise the structure of an argument or generate an example. Students are then directed back to the original reading to check accuracy, recover nuance and build their own understanding.

This distinction between orientation and close reading is especially useful in first year. Many students are still learning how to read academically. Simply telling them not to use AI may ignore the genuine difficulty of entering university-level study. Telling them to use AI without guidance may allow the tool to flatten complexity or replace the reading altogether. A more educative approach is to show students what AI can assist with and what it cannot do for them.

A small piece of guidance beside a reading can do a lot of work. It might say: 

• use AI to identify the main sections of the article but return to the article to understand the author’s argument
• use AI to clarify unfamiliar terms but check the explanation against the course material
• use AI to generate an example but decide whether that example reflects the concept being taught.

That is AI literacy in course design: a set of repeated, explicit learning moments where students are shown how to approach AI in relation to the work they are already doing.

Make AI something that students question

In the course focused on how students learn, students encounter AI as something to question, not simply something to use. They might examine an AI-generated explanation of a learning theory and compare it with course concepts and readings. The task is to ask whether it is accurate, what it oversimplifies or leaves out, and what a teacher would need to know before relying on it.

This positions AI as a prompt for judgement. They still need to read, interpret, compare and revise. The AI output becomes useful only when students bring their disciplinary understanding to it.

Be explicit about responsibility

My approach is informed by the ethical augmentation framework, which centres human judgement, relational practice and ownership of learning. In practical terms, this means AI is a support for learning, not a substitute for thinking. Students may use AI where it can reduce unnecessary cognitive load, clarify a starting point or support reflection, but they remain responsible for interpretation, accuracy, ethical decision-making and professional judgement.

The relational aspect is easy to overlook. Trust, clarity and expectations shape students’ use of AI. If they are worried that any use of AI will be treated as misconduct, they may hide it rather than learn from it. If they are told only that AI is allowed, they may assume that permission equals educational value. Neither position supports mature academic practice.

Students need explicit guidance. They need to understand why AI might help clarify a reading but cannot replace the act of reading. They need language for describing different uses, such as clarification, brainstorming, feedback, editing or verification. These are not the same, and students need to learn the difference.

Design from the judgement students need

In teacher education, responsible AI use is connected to professional formation. A pre-service teacher who uses AI to generate a learning activity still needs to judge whether it is developmentally appropriate, inclusive, evidence-informed and responsive to learners. A future teacher cannot delegate professional judgement to a tool.

Other disciplines will need different examples, but the design principle is transferable. Start with the judgement students need to develop. Look for the places where they already interpret, question, compare, justify, apply or reflect. These are the points where AI literacy can be embedded without turning it into an add-on.

Educators do not need to redesign everything at once. Add guidance beside a reading. Build a tutorial activity where students critique an AI-generated response against course concepts. Ask students to explain why they accepted, rejected or revised a suggestion. Make space for students to discuss not only whether AI was used, but whether that use supported learning.

If a course claims to value critical thinking, professional judgement or ethical practice, AI guidance should not sit separately from those commitments. Judgement cannot be bolted on. Students need to practise it in context, with support, feedback and clear expectations.

Nicole Brownlie is a lecturer in teacher education at the University of Southern Queensland.

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.