Questions

No. Academic Supervisor will critique, question, challenge and teach. It will not draft, rewrite, paraphrase or “improve” a sentence of your research — not once, not for a heading, not for an abstract.

This is not a policy we ask you to trust. The AI Supervisor can only reply in a fixed critique format — strengths, weaknesses, omissions, questions — and any reply that comes back as ordinary prose is rejected before you ever see it. There is no autocomplete, no rewrite button and no way to insert AI Supervisor text into your document.

Completely. A ghostwriter hands you sentences you did not think of and cannot defend. Academic Supervisor hands you the questions your examiner will ask, and leaves you to answer them.

The practical difference shows up in a viva. Every weakness the AI Supervisor raises arrives as a question you have to answer yourself — “which studies did you search, and what do they leave open?” — so by the time you finish, you have already defended the work once.

That is your institution's decision, and you should tell your supervisor you are using it — we would rather you declared it than hid it.

What we can say is that the thing institutions prohibit is submitting work you did not write. Nothing this tool produces can be submitted: it returns critique, not content. Most academic-integrity policies distinguish between a tool that generates text for you and one that gives feedback on text you wrote. This is firmly the second.

Yes, and this is the part detection software cannot give you. Every time you accept a section, Academic Supervisor keeps a snapshot. Your project carries the whole history of how the writing developed — early drafts, the critiques you received, the changes you made in response, and the concerns you considered and deliberately kept your own view on.

That record is evidence of process. A finished document proves nothing about who wrote it; twenty revisions with critique in between is a much harder thing to fake.

Never, deliberately. AI systems invent references that look perfect — right-sounding author, plausible journal, wrong or non-existent paper — and a fabricated citation in your bibliography is your problem at the viva, not the model's.

You record what you have actually read in the Literature table. The AI Supervisor reviews your reading for synthesis, coverage and gaps. It will tell you that a claim is unsupported. It will not tell you what to cite.

No — and it is built so that it cannot. It does the part of supervision that does not need you: the fifth reading of a problem statement, the obvious methodological mismatch, the question the student could have anticipated. It arrives at your desk having already been asked.

Judgement about the field, the standard expected, and whether this student is ready to proceed stays with the supervisor. The AI Supervisor has no view on any of that.

Yes. Supervisors get their own workspace listing the students assigned to them, with each student's work and a comment box on every section, on the Alignment Matrix and on the Literature table.

Access is read-only, always. A supervisor cannot edit a student's writing, cannot run the AI Supervisor on their behalf, and cannot make the work anything other than the student's own.

The student, per project. Your institution assigns the supervisor; the student then chooses which of their projects that supervisor may open, and can stop sharing at any time. Comments already made stay visible to the student.

By default a supervisor sees the writing and the plan, but not the student's AI Supervisor history — so a student can admit to the AI Supervisor that they are stuck without it becoming part of their assessment. A student can grant that wider access if they want to.

Every institution currently works from the centrally approved structures and rubrics — the same ones described on the governance page. Institution-specific structures are the next thing we are building, and pilot departments help decide what they look like. If your faculty has a required proposal format, talk to us.

An administration screen: your staff and students on one roster, supervisor assignments made and ended in one place, and your own institutional space separate from every other. Nobody outside your institution sees your roster.

A research proposal, a thesis or dissertation, a journal article, a paper review or a term paper. Each comes with the standard structure for that document type, and sections unlock in order — you cannot write a methodology before you have settled a research question, because that is the mistake that costs a year.

Reading, writing, saving, accepting and exporting are free and always will be — you are never charged to work on your own document. Only asking the AI Supervisor costs credits.

Credits are priced by how much text the AI Supervisor has to read, so a short section costs less than a long one, and the button shows you the exact price before you click it. New accounts start with credits to try it. If a response comes back malformed, you are not charged.

Whichever your department requires — you set the style on each project, and the AI Supervisor holds your Literature entries to it. It checks your consistency; it does not supply the references.

You, and any supervisor you have chosen to share a project with. That is the complete list. We cannot open your writing — support can see technical records of what happened and when, and nothing else.

Your writing is not used to train any AI model.

It stays yours. Export your writing, your Alignment Matrix and your Literature at any time — the export contains your work only, never a word of AI Supervisor text. To have your account and everything in it deleted, email support; what we keep and for how long is set out in the privacy notice.