AI use academic misconduct

Short answer

There is no single rule for England and Wales. Whether using AI for planning, coding or brainstorming counts as academic misconduct depends entirely on your institution’s regulations, the specific assessment brief, and above all whether the work you submit is genuinely your own and properly acknowledged. Using AI to spark ideas or plan is often permitted; using it to generate the substance of the work you then pass off as yours is usually misconduct.

Why there is no universal answer

Academic misconduct is not defined by general law. It is governed by the contract between you and your institution, made up of the student regulations, the academic integrity or assessment misconduct policy, and the rules for each individual module or assessment. Because each university, college or awarding body sets its own rules, the same use of AI can be entirely legitimate on one course and a serious offence on another. As UCL’s guidance puts it, integrity is fundamentally about fairness, so whether a use is fair depends on the category and instructions of the particular assessment. There is no fixed national list of permitted and prohibited AI uses.

This means the decisive questions are always: what does my institution’s policy say, what does the assessment brief say, and have I complied with any disclosure requirement?

The general distinction most institutions draw

Across mainstream university guidance a consistent pattern emerges.

Using AI at an early stage to overcome writer’s block, generate ideas, suggest topics to investigate, provide examples, or help you develop your own plan or structure is commonly treated as acceptable, provided you engage critically with the output and it does not replace your own thinking. Newcastle University, for example, expressly describes using AI at the planning stage to generate ideas as acceptable use where you use the suggestions critically. Westminster lists brainstorming ideas and developing a draft plan as permissible uses subject to acknowledgement.

By contrast, submitting work that has been generated wholly or in part by an AI tool as if it were your own is generally academic misconduct. Newcastle states plainly that work generated wholly or in part by an AI tool, or by another person, would be misconduct. The core principle running through the sector is that the work you submit for assessment must be your own.

So the safest way to understand the boundary is this. AI that helps you think is usually fine. AI that does the thinking, writing or producing for you, which you then present as your own effort, is usually not.

Where planning, coding and brainstorming each fall

Brainstorming and idea generation tend to sit on the safer side, because the intellectual output remains yours. The risk arises if you lift AI-generated phrasing or arguments verbatim, or if the assessment is specifically testing your ability to generate ideas independently.

Planning is similar. Using AI to help structure an essay or project is often allowed, but if the assessment is testing your ability to plan or design a methodology, having AI do it may undermine the learning outcome being assessed and cross into misconduct.

Coding is the most variable. On some computing modules, using AI coding assistants mirrors professional practice and is encouraged or permitted with acknowledgement. On others, particularly assessments designed to test whether you personally can write and understand code, submitting AI-generated code as your own is a clear breach. Copying AI-produced code without understanding it also creates a real risk in any viva or follow-up questioning, where you may be asked to explain your own work.

The acknowledgement requirement

Even where AI use is permitted, many institutions require you to declare it, and failing to declare it can itself be misconduct. Westminster, for instance, requires a statement such as “In this assessment, I used [tool] to [explain how]”. The University of Westminster treats undisclosed use as potentially constituting academic misconduct even where the use itself would have been allowed if declared.

This is a crucial and often overlooked trap. The offence is frequently not the AI use itself but the non-disclosure. If your institution has a disclosure format, use it precisely.

How different facts change the outcome

The answer turns heavily on details you have not given, so the position could shift depending on the following.

Whether your assessment brief expressly permits, restricts, or bans AI. A blanket ban means even brainstorming could be a breach; an open policy may permit substantial use with acknowledgement.

What the assessment is testing. If the very skill being assessed is the one you delegated to AI, the risk is far higher.

How much of the final submitted work originates from AI, and whether you have transformed it through your own understanding and expression, or reproduced it.

Whether you disclosed the use in the required form.

Whether you are at a university, a further education college, or under a specific awarding body, each of which has its own rules.

Practical next steps

1. Read the exact wording of your institution’s academic integrity or assessment misconduct policy, and the specific assessment brief, before you rely on any AI tool. These override any general guidance.

2. Look for an AI-use statement or declaration requirement, and follow its wording exactly if AI use is permitted.

3. If the position is unclear, ask the module leader or academic in writing before submitting, and keep the reply. Written confirmation is strong protection if a query later arises.

4. Keep a record of how you used AI, including your prompts and what you did with the output, so you can demonstrate that the final work reflects your own understanding.

5. For coding especially, make sure you can independently explain and justify every part of what you submit, in case of a viva or follow-up questions.

Bottom line

Using AI for planning, coding or brainstorming is not inherently academic misconduct, and is often expressly permitted for early-stage idea generation and structuring. It becomes misconduct where it replaces your own work, where it produces the substance you then present as your own, where it undermines the specific skill being assessed, or where your institution requires disclosure and you fail to give it. Because the rules are set by your particular institution and assessment, the decisive step is to check those specific rules rather than rely on any general principle.