AI for dissertation writing that works with you, not for you

Who does what

Your dissertation is marked as your work, so the split needs to be clear from day one. Here is the one we suggest:

What it will never do is attribute experiences, data or findings to you that you haven’t given it. If your project involves interviews, a survey, a placement or lab work, that material comes from you.

What it looks like while you work

On screen, you talk to the guide on one side while your document grows on the other, and you can edit it at any point or undo changes just as you would in Word. At the start, you pick the type of paper or attach your module handbook, and the suggested outline then appears in the document. Halfway through, you have the chapter being drafted next to your source library. At the end, the download panel has the title page filled in and the referencing style you chose.

A typical dissertation structure

Every department sets its own in the handbook, and an empirical project looks different from a desk-based or literature-based one. This is the most common layout for an empirical dissertation:

In a literature-based dissertation, the methodology describes your search (databases, keywords, inclusion criteria) and the findings are your synthesis of the studies. The guide to writing a dissertation walks through both types.

Referencing styles you are likely to meet

Many UK departments use a version of Harvard, psychology and education usually ask for APA 7, health sciences often use Vancouver, engineering uses IEEE and some humanities departments prefer MHRA or Chicago. Always follow your handbook: departmental versions of Harvard differ in small ways.

In Aithor.io you pick the style in the paper’s Sources panel and can change it at any time: 10,851 citation styles (APA 7, MLA, Chicago, Harvard, IEEE, Vancouver, AMA, ABNT…). Before you submit, compare the first few references with your department’s style guide.

Showing how you wrote it

More supervisors now want to know how a dissertation was produced, not just read the result. The Process report sums it up in one panel. Shows how much you wrote, who wrote each part (typed, pasted, edited from the guide or from the guide) and how it was done (days, messages and edits). It measures what happened in the editor: it does not certify that no AI was used.

Don’t lean on AI detectors either, whether to accuse or to reassure yourself: a study that tested 14 detection tools, including two commercial systems widely used by universities, concluded they were neither accurate nor reliable ^[3].

If your university asks you to declare AI use, be specific: which tool, for which parts and how you checked what it suggested. The Russell Group principles say universities should make clear where generative AI is inappropriate and help students acknowledge their use where necessary ^[1], and the QAA has resources on keeping assessment sound with generative AI ^[2]. To see how rules differ between institutions, read what UK and US university policies say about AI in dissertations. There is a template in our guide on how to disclose AI use in an assignment.

Start for free

Preparing for your viva or presentation

Defending a dissertation means explaining choices: why this question, this method, these sources. Defense mode: asks you about your paper one question at a time and tells you what to review. That way you find the weak spots before your examiners do.

For your slides, you can send the dissertation straight to Study and turn it into a presentation: Editable PowerPoint with speaker notes, PDF and 15 designs. The full process is in viva and presentation practice.

Planning your time

If you start about three months before the deadline, work in this order and send your supervisor something at the end of each stage:

1. Weeks 1–2: read the handbook, choose a topic and narrow the question. Send a working title and your aims.
2. Weeks 3–6: search and read, apply for ethics approval if you need it, and draft the literature review. Send the outline and the review.
3. Weeks 7–10: methodology, data collection (or your search protocol), findings and discussion. Send a full draft.
4. Weeks 11–12: introduction, conclusion and abstract, then formatting, references, your AI declaration and viva practice.

The introduction and abstract come near the end on purpose: it is much easier to introduce a project that already exists. For a week-by-week version with a downloadable template, see our dissertation timeline. For a ready-made skeleton in Word, use the dissertation outline template.

Aithor.io compared with a general AI chat

A general-purpose chatbot can help with a one-off question about your dissertation. What it doesn’t do by default is hold a long project together with its outline, sources and history. The table sets it out point by point.