AI and Academic Integrity: A Guide for Australian Teachers

Academic integrity has always been a challenge in schools. AI tools have made it significantly more complex. Students can now generate written work that is coherent, well-structured, and difficult to distinguish from their own writing — in seconds, for free.

This guide addresses the practical questions Australian teachers are facing: how to update assessment design, how to respond when AI use is suspected, how to teach students to use AI responsibly, and what the current guidance from Australian education departments says.

---

Why Detection Is Not the Answer

The instinctive response to AI-generated student work is to look for a detection tool. Several AI detection tools exist — Turnitin has integrated AI detection into its platform, and standalone tools like GPTZero are available. However, relying on detection tools as the primary response to AI and academic integrity is problematic for several reasons.

Detection tools are unreliable. AI detection tools produce false positives — flagging human-written work as AI-generated — at rates that are unacceptably high for making consequential decisions about students. Research has found that detection tools are particularly unreliable for students who are non-native English speakers, whose writing patterns can resemble AI output.

The technology is moving faster than the detection. AI tools are improving rapidly, and detection tools are consistently behind. Work generated by the most capable current AI models is increasingly difficult to detect reliably.

Detection is adversarial. A detection-focused approach positions teachers and students as adversaries, which is not conducive to the learning relationship. The more productive approach is to design assessments that make AI use less effective, and to educate students about when and how AI use is appropriate.

---

Assessment Design: The Most Effective Response

The most effective response to AI and academic integrity is assessment design. Assessments that require students to demonstrate specific knowledge of their own learning context, that include oral components, or that are completed under supervised conditions are significantly harder to complete using AI tools.

Strategies for AI-Resistant Assessment Design

Require personal and contextual knowledge. Tasks that ask students to connect content to their own experiences, their local community, or their specific learning journey are harder to complete with AI. "Analyse the themes in the novel we studied" is easier to AI-generate than "Explain how the themes in the novel connect to something you have personally experienced or observed."

Include process documentation. Requiring students to submit drafts, planning notes, or a reflection on their writing process makes it harder to submit AI-generated work without evidence of genuine engagement.

Use oral components. An oral presentation, a viva voce, or a brief oral defence of written work requires students to demonstrate understanding that cannot be delegated to an AI tool.

Design for specificity. Tasks that require students to respond to specific stimuli — a particular text studied in class, a specific local issue, a recent event — are harder to AI-generate than generic tasks.

Use supervised assessment. For high-stakes tasks, supervised conditions remain the most reliable way to ensure authentic student work.

---

Current Australian Department Guidance

Each Australian state and territory education department has developed guidance on AI and academic integrity. The common themes are:

Schools should update their academic integrity policies to explicitly address AI-generated work. Most existing policies were written before AI tools were widely accessible.

Context matters. Whether AI use is appropriate depends on the task and the explicit instructions given to students. Using AI to brainstorm ideas for an essay may be appropriate; using AI to write the essay and submitting it as your own work is not.

Students need to be educated, not just policed. Departments consistently emphasise that the goal is to develop students' AI literacy — their ability to use AI tools appropriately and critically.

Teachers should not make academic integrity findings based on detection tool output alone. The unreliability of detection tools means that any finding needs to be based on a broader assessment of the evidence.

---

Teaching Students to Use AI Responsibly

What to Teach

What AI tools are and how they work. Students who understand that AI tools generate text by predicting likely word sequences — rather than by thinking, knowing, or understanding — are better equipped to evaluate AI output critically.

When AI use is appropriate. AI tools are appropriate for some tasks (brainstorming, getting feedback on a draft, understanding a concept) and not for others (submitting AI-generated work as your own). Students need explicit guidance about the distinction.

How to use AI tools to support learning. Using AI to explain a concept you do not understand, to get feedback on a draft you have written, or to generate ideas you then evaluate and develop is a legitimate and valuable use of AI.

Critical evaluation of AI output. AI tools produce errors, biases, and oversimplifications. Students who learn to evaluate AI output critically develop skills that are valuable beyond the academic integrity context.

Classroom Activities

AI output analysis. Give students an AI-generated response to a question and ask them to identify what is accurate, what is inaccurate, what is missing, and what could be improved.

Prompt engineering. Have students experiment with writing prompts for an AI tool and compare the outputs. This builds understanding of how AI tools work.

Ethical discussion. Use discussion activities to explore the ethical dimensions of AI use — when is it fair to use AI? What does it mean to do your own work? How should AI use be disclosed?

---

Responding to Suspected AI Use

When a teacher suspects that a student has submitted AI-generated work, the appropriate response depends on the school's academic integrity policy. General principles that apply across most contexts:

Start with a conversation. Before making any finding, have a conversation with the student about their work. Ask them to explain their argument, describe their research process, or talk through their thinking.

Document your observations. Note the specific features of the work that raised your concern, and document the conversation with the student.

Follow your school's process. Academic integrity findings should follow the school's established process, which typically involves the student, the teacher, and a member of school leadership.

Be cautious about detection tool output. If you have used a detection tool, treat its output as one piece of context, not as evidence.