The public conversation about AI in Australian education has been dominated by two themes: students cheating with ChatGPT, and whether AI will replace teachers. Neither captures what’s actually happening in Australian classrooms, which is considerably more interesting and more complicated than the headlines suggest.

The Ban-Then-Unbann Cycle

When ChatGPT launched, most Australian state education departments panicked and banned it. That was understandable if short-sighted. Within twelve months, most had reversed course and were developing guidance for AI use in education.

The current state varies by jurisdiction. New South Wales has been the most proactive, with the Department of Education providing detailed guidance for teachers on integrating AI into classroom practice. Victoria has developed an AI in education framework that’s reasonably comprehensive. Queensland and South Australia are somewhere in between.

What’s consistent across all jurisdictions: the recognition that banning AI isn’t viable and that students need to learn to use it effectively rather than be prevented from accessing it.

What Teachers Are Actually Doing

I’ve visited over a dozen Australian schools in the past year, talking to teachers about their real AI practices. Here’s what I found.

Lesson planning and resource creation. This is the most common use. Teachers are using AI to generate lesson plan drafts, create differentiated worksheets, and develop assessment rubrics. Most teachers I spoke with said AI saves them two to five hours per week on administrative tasks, freeing time for actual teaching.

Personalised learning support. Some schools are using AI tools to provide personalised practice and feedback for students. In maths, platforms that adapt difficulty based on student performance are widespread. In English, AI writing feedback tools are gaining traction, though teachers are careful to position them as supplements to human feedback rather than replacements.

Teaching about AI. A growing number of schools are explicitly teaching students about AI: how it works, what it can and can’t do, and how to use it critically. This isn’t limited to computing classes. English teachers are teaching students to evaluate AI-generated text. Science teachers are using AI to demonstrate pattern recognition and data analysis.

Assessment redesign. This is the most significant and challenging area. Teachers are redesigning assessments to remain meaningful in a world where students have AI access. Approaches include process-based assessment (evaluating how students approach problems, not just final answers), oral examinations, in-class supervised tasks, and portfolio-based assessment that demonstrates learning over time.

The Equity Problem

Here’s the issue that gets too little attention. AI access in Australian schools is deeply unequal.

Well-funded private schools and government schools in affluent areas have invested in AI tools, trained their teachers, and developed integration strategies. Schools in lower socioeconomic areas are still struggling with basic IT infrastructure, let alone AI integration.

The Digital Education Revolution was supposed to address the digital divide. It partially did, but the AI divide is a new dimension that existing programs don’t adequately address.

Students in well-resourced schools are learning to use AI as a thinking partner and productivity tool. Students in under-resourced schools aren’t getting that preparation. When both groups enter the workforce, the advantage will compound.

State governments need to address this actively. That means funding for AI tools in disadvantaged schools, professional development for teachers in those schools, and explicit programs to ensure AI literacy isn’t another dimension of educational inequality.

The Assessment Integrity Debate

Let’s talk about cheating, because it hasn’t gone away.

AI-generated text is increasingly difficult to detect. The detection tools that schools initially adopted have proven unreliable, generating false positives that unfairly penalise students and false negatives that miss AI-generated work.

Most schools have stepped back from relying on detection tools and instead moved toward two strategies. First, redesigning assessments to be AI-resistant (though this is an arms race). Second, developing academic integrity approaches that focus on the process of learning rather than just the product.

The second approach is more sustainable. If a student can demonstrate their understanding through conversation, progressive drafts, and in-class work, the question of whether their take-home essay was AI-assisted becomes less important. The learning happened regardless.

Some educators argue that using AI to assist with assignments is no different from using a calculator or spellchecker. Others maintain that there’s a fundamental difference between tools that help you execute your thinking and tools that do the thinking for you. Both positions have merit, and the debate is genuinely unresolved.

What Students Think

I’ve talked to students too. Their perspective is pragmatic.

They’re going to use AI in their future workplaces. They know this. Learning to use it effectively seems obvious to them. The idea that school should pretend AI doesn’t exist strikes most students as absurd.

They also recognise the risks. Several students expressed concern about becoming dependent on AI for tasks they should learn to do themselves. They understand the difference between using AI as a tool and using it as a crutch, even if they don’t always make the right choice.

The most thoughtful students I spoke with wanted schools to teach them AI literacy explicitly: how to evaluate AI outputs, when to trust them, when to question them, and how to combine AI assistance with their own thinking. That seems like a reasonable ask.

Where This Goes

AI in Australian education is heading toward integration, not prohibition. The schools that are doing it well are treating AI as a pedagogical tool that requires new teaching approaches, not just a technology problem to manage.

The challenge is speed. AI capabilities are advancing faster than curriculum, teacher training, and assessment practices can adapt. Schools that figure out how to learn and adapt continuously, rather than implementing fixed AI policies, will serve their students best.

Australian education has navigated technology transitions before. This one is faster and more disruptive than previous waves, but the fundamental approach, focusing on learning outcomes rather than technology anxiety, remains sound.