AI and Marking

Artificial intelligence is already becoming part of everyday teaching. Like calculators or search engines before it, the question is less about whether teachers should use it and more about thoughtful and ethical use.

The latest OECD Education and Skills Teaching and Learning International Survey (TALIS 2024) found that around one in three teachers across OECD education systems are now using AI in their work. Of those teachers using AI, around 26% reported using it to assess or grade student work.

In New Zealand, the numbers are even higher. Around 69% of Year 7–10 teachers reported using AI in their work compared with the OECD average of 36%. However, only around 12–14% of New Zealand teachers reported using AI to assess or mark student work.

From my experience working with teachers and students, I have found that AI can be incredibly useful for reducing repetitive workloads. Marking and feedback take a huge amount of my time, and AI can help identify patterns, generate draft feedback, highlight misconceptions, or support consistency across large groups of student work. Used well, it can free me up to spend more time doing the human side of education, conferencing with students, building relationships, adapting learning, and responding to individual needs.

AI should not replace teacher judgment, especially in areas that require nuance, cultural understanding, creativity, or relational knowledge of the student. A teacher knows when a student has taken a genuine intellectual risk, shown growth in confidence, or demonstrated understanding in ways that do not always fit neatly into a rubric. AI does not truly understand context, emotion, identity, or the lived experiences students bring into the classroom.

One of my concerns is that if schools move toward fully automated marking systems for efficiency, we risk reducing learning to what is easiest for a machine to measure. Good education is often messy, relational, and deeply human. The feedback process itself is also part of the relationship between teacher and student. Students need to feel seen, understood, and supported, not just processed.

That said, AI can play a valuable supporting role when teachers remain accountable for the final judgment. In my own practice, I see AI as a thinking partner rather than a replacement for professional expertise. The most effective use is usually when AI helps teachers notice trends, draft possible feedback, or save time on administration, while the teacher still verifies, adapts, and personalises the response.

So yes, teachers using AI can be a good idea, but only if we keep humans firmly at the centre of the process. The goal should not be to remove teachers from the learning process. It should be to remove some of the barriers that stop teachers from spending meaningful time with students.