Every exam body I talk to feels the same squeeze. Governments want innovation. Schools want recognition for richer work. Generative AI is crashing into the system from all sides. And yet, when results day comes, the only thing that really matters is whether the grades stand up in the media and in court.
AI is arriving in education as if it were a cure‑all for workload and consistency. Sales decks promise tools that “judge writing like teachers”, “skip marking altogether”, and “cut workload by 90 per cent”. It is an attractive story in a system under pressure. But if we listen carefully to the people building these systems – and to the artists responding to them – a different story emerges.
In 1792, revolutionaries in Paris abolished a king, Americans calmly re‑elected a president, and Cambridge quietly invented something that still shapes millions of lives every year: exam marking. While politics and industry were being rebuilt in public, assessment was being rebuilt on paper. Two centuries later, we are still living inside that decision – and only now starting to see its limits.
Over this series, we’ve tried to do three things: explain why UK copyright policy on AI is shifting, argue that student work should be treated as creative property rather than free fuel, and sketch when – if ever – training on that work might be legitimate.
Every exam body I talk to feels the same squeeze. Governments want innovation. Schools want recognition for richer work. Generative AI is crashing into the system from all sides. And yet, when results day comes, the only thing that really matters is whether the grades stand up in the media and in court.
AI is arriving in education as if it were a cure‑all for workload and consistency. Sales decks promise tools that “judge writing like teachers”, “skip marking altogether”, and “cut workload by 90 per cent”. It is an attractive story in a system under pressure. But if we listen carefully to the people building these systems – and to the artists responding to them – a different story emerges.
In 1792, revolutionaries in Paris abolished a king, Americans calmly re‑elected a president, and Cambridge quietly invented something that still shapes millions of lives every year: exam marking. While politics and industry were being rebuilt in public, assessment was being rebuilt on paper. Two centuries later, we are still living inside that decision – and only now starting to see its limits.
Over this series, we’ve tried to do three things: explain why UK copyright policy on AI is shifting, argue that student work should be treated as creative property rather than free fuel, and sketch when – if ever – training on that work might be legitimate.
So far in this series, we’ve argued that student work is creative property, not just “data”, and that assessment use and training use are not the same thing. That naturally leads to a harder question: is it ever acceptable to train models on student work – and if so, on what terms?
When people talk about AI in assessment, the conversation usually goes straight to marking accuracy, bias or workload. Underneath all of that sits a quieter question that is just as important: who actually owns the work being processed – and what does that ownership mean once AI is involved?
For a while, it looked like AI developers might get a broad “free pass” to train on almost anything they could scrape. That idea has now been quietly parked. The UK government’s latest report on copyright and artificial intelligence signals a very different direction – one that matters a lot if you’re responsible for student work and assessment.
Every year in England, some things in education feel almost guaranteed. Exam season will arrive on schedule, bringing with it the familiar mix of anxiety, hope and hard work in schools up and down the country. Ofqual will emphasise that our qualifications system is robustly designed, closely regulated and delivering grades that are fair and can be trusted. And Dennis Sherwood will publish fresh analysis arguing – drawing largely on Ofqual’s own technical reports – that neither the level of fairness nor the level of trust we assume is quite what it seems.
The latest OECD Reimagining Teaching in an Accelerating World report makes one thing very clear: teaching and assessment can’t stay as they are. As AI reshapes what students can do with a few prompts, the things that are easiest to test have become the easiest to automate. Systems everywhere are being pushed to value richer learning, trust professionals more, and treat AI as an educational ally rather than a threat.