Yes, students can use AI tools during coursework, NEAs (non-exam assessments), and other internal assessments where internet access is permitted — but only if they properly reference the AI use, keep a non-editable record (like a screenshot) of what was generated and how it was used, and ensure the final submitted work reflects their own understanding. This is set out in guidance from JCQ (the Joint Council for Qualifications), which represents all major UK exam boards. Undisclosed or unreferenced AI use can be treated as malpractice, and a teacher who can’t authenticate the work as genuinely the student’s own is required to submit a mark of zero.
If your child has coursework, a GCSE NEA, or any assessment that allows internet access during preparation, “can I use AI for this” is a fair, non-obvious question — and the honest answer is more nuanced than a simple yes or no.
Who Sets These Rules, and Where Do They Apply?
The rules come from JCQ, the body representing AQA, Edexcel/Pearson, OCR, WJEC, and the other major UK exam boards together. Their guidance on AI use in assessments was first published in 2023 and has been updated since, most recently with clearer real-world examples of what counts as misuse.
Crucially, these rules apply specifically to coursework, NEAs, and other internal assessments where internet access is permitted during preparation — not to exams themselves. The vast majority of GCSE and A-Level exams are still sat under close supervision with no internet or AI access at all; the AI guidance is entirely about the research and production stage of coursework-style assessments.
What’s Actually Allowed
- Using AI to research a topic, in the same way a student might use a search engine or textbook, as long as the sources (including the AI tool itself) are properly referenced.
- Using AI to help understand a concept before producing original written work based on that understanding.
- Using AI-generated content as a starting point, provided this is clearly disclosed and the final submission reflects the student’s own analysis and writing, not a lightly edited copy of AI output.
What Counts as Malpractice
- Submitting AI-generated content as the student’s own work, without disclosure.
- Failing to keep the required record. Students must retain the question(s) asked and the AI-generated content in a non-editable format, such as a screenshot, along with a brief explanation of how it was used — and this needs to be available for the teacher to review.
- Work the teacher can’t authenticate as genuinely the student’s own. If a teacher has real concerns and can’t sign the authentication statement, that piece of work is not accepted for assessment, and a mark of zero is submitted for it — a serious consequence that’s easy to avoid with proper disclosure from the start.
JCQ has published real examples of AI-related malpractice cases specifically to help schools and students understand where the line actually sits, since “properly referenced” can otherwise feel vague.
A Practical Way to Think About It
The underlying principle across all of this is straightforward, even if the specific rules feel technical: a student can’t be credited for work that isn’t genuinely their own. Used well, AI is treated similarly to any other research or reference tool — helpful for understanding and exploring a topic, but not a substitute for the student’s own analysis, writing, and final understanding. Used to shortcut the actual work, it creates real risk, both academically (a grade zero on that piece) and in terms of the underlying learning gap it leaves behind for the actual exam, which has no AI access at all.
What Parents Should Actually Do
- Ask your child’s school what their specific AI policy is. JCQ sets the baseline rules, but individual schools and departments may have additional, more specific guidance for their students.
- Encourage disclosure, not avoidance. A student who used AI and properly referenced it is in a completely different position from one who used it and didn’t — the AI use itself isn’t the risk, the lack of disclosure is.
- Keep the exam in mind. Even coursework that’s completed with fully compliant AI use doesn’t prepare a student for the closed-book exam room, where none of these tools are available. If AI use during coursework is replacing genuine understanding rather than supporting it, that gap will surface at exam time regardless of how the coursework mark turned out.
Want coursework support that builds real understanding your child can rely on, exam room included? Book a free diagnostic session and we’ll help your child use every resource properly, without the shortcuts that catch up with them later. Explore our full exam preparation support or UK curriculum tutoring.
Q: Can students use ChatGPT or similar AI tools for GCSE coursework?
Yes, for coursework, NEAs, and other assessments where internet access is permitted during preparation, provided the use is properly referenced and a record of the AI-generated content is kept, in line with JCQ guidance.
Q: What happens if a student doesn’t disclose AI use in coursework?
This can be treated as malpractice. If a teacher cannot authenticate the work as genuinely the student’s own, the work is not accepted for assessment, and a mark of zero is submitted for that piece.
Q: Are these AI rules different for exams versus coursework?
Yes. The JCQ AI guidance applies specifically to coursework, NEAs, and other internal assessments where internet access is allowed during preparation. The vast majority of exams remain closed-book and closely supervised, with no internet or AI access permitted at all.
Q: Do all UK exam boards follow the same AI rules?
Yes. JCQ represents the major exam boards — including AQA, Edexcel/Pearson, OCR, and WJEC — together, so the core guidance on AI use in assessments applies consistently across them, though individual schools may add their own specific policies on top.