Academic integrity · 9 minute read
Is using AI on your dissertation cheating?
Short answer: it depends on what you ask it to do. Using AI to ghostwrite chapters, fabricate citations, or evade detection is dishonest. Using AI for reflection, alignment checks, and structured feedback is a different question — and one worth examining carefully.
The short version
- Most AI policy violations involve outsourcing authorship, not using AI for reflection or process support.
- The key test: can you defend every claim, source, and decision to your committee without the AI?
- Integrity-first tools are designed to refuse ghostwriting and support scholarly judgment.
§ 1 · Policy landscape
What institutional policy actually says
Most doctoral programs have not yet adopted specific AI policies. Some explicitly permit AI for brainstorming, checking alignment, or summarizing your own notes. Others prohibit any AI-generated content in submitted work. A growing number require disclosure.
The absence of a policy does not mean anything is permitted. The default is that submitted work must reflect your own intellectual effort. If you are unsure, ask your chair or doctoral office directly — and document the response.
- Does your program have a written AI policy? Where is it published?
- Does 'your own work' in your institution's academic integrity policy include AI-assisted drafting?
- What disclosure format does your program require for AI use, if any?
§ 2 · The real line
What actually constitutes academic dishonesty
Using AI to produce ready-to-submit dissertation prose you then claim as your own is academic fraud — the same category as contract cheating. The same applies to using AI to paraphrase your work to evade detection software, to fabricate citations, or to produce interpretations you cannot defend.
What is generally not considered dishonest: using AI to generate questions that challenge your reasoning, to build a reading list you then independently evaluate, to check your own writing for clarity, or to run an alignment check on elements you authored.
- Ghostwriting chapters or sections: dishonest.
- Fabricating citations or quotations: dishonest.
- Paraphrasing to evade AI detection: dishonest.
- Using AI-generated prose without disclosure: dishonest.
- Asking AI to question your reasoning: permitted in most programs.
- Running an alignment check on your own draft: permitted in most programs.
- Using AI to check grammar with your explicit edits: permitted in most programs.
§ 3 · Integrity-first design
How integrity-first tools differ from writing AI
Dissertation Coach is built to refuse ghostwriting. It will not write chapters, generate ready-to-submit prose, or paraphrase your work to evade detection. What it does: ask Socratic questions, flag alignment gaps, suggest next actions, and support your scholarly judgment.
The distinction matters for policy compliance: a tool that asks 'What would falsify your central claim?' is not producing your dissertation — it is helping you test your own reasoning.
§ 4 · Disclosure
When and how to disclose AI use
Even where disclosure is not required, developing a habit of transparency protects you. Keep a brief log of when you used AI, for what purpose, and what you retained or rejected. This makes disclosure easy and audit-proof if your institution later requires it.