Academic Integrity & Ethical AI Use
This is the most important lesson in the entire course. Read it twice.
Academic integrity is not a bureaucratic detail. It is the foundation of your entire degree, your transcripts, your references, and the trust that future employers and graduate programs place in your credentials. AI does not change that — it raises the stakes. The students who get this right will look back on AI as the most powerful research tool of their generation. The students who get it wrong will look back on it as the reason they failed a course or were expelled.
This lesson is about how to use AI in a way that is honest, defensible, and ultimately makes you a better researcher.
What You'll Learn
- The difference between using AI as a tool and using AI as a ghostwriter
- What your university's policy probably says (and how to find out for sure)
- When and how to disclose AI use in academic work
- The five red lines you must never cross
Tool vs Ghostwriter: The Core Distinction
A calculator is a tool. You can use a calculator to do arithmetic on a problem set, and nobody accuses you of cheating, because the cognitive work — setting up the problem, choosing the method, interpreting the result — was yours.
AI used as a tool looks like this:
- Asking Claude to explain a confusing concept from a paper you already read
- Using Perplexity to find papers on a topic you have already started exploring
- Asking ChatGPT to suggest counterarguments to a thesis you have already developed
- Using Grammarly to fix grammar in a draft you wrote
- Asking an AI to summarize a paper to decide if it is worth reading in full
AI used as a ghostwriter looks like this:
- Pasting an assignment prompt into ChatGPT and submitting the output
- Asking AI to "write a 2000-word essay on X" and editing the result lightly
- Generating fake citations to support claims you did not actually research
- Having AI rewrite a real source's text and submitting it as your analysis
The difference is who is doing the thinking. If the structure of the argument, the choice of evidence, and the conclusions are yours, AI is a tool. If they are the AI's, AI is a ghostwriter — and that is academic misconduct everywhere.
Find Out Your University's Actual Policy
Do not guess. Do not ask a friend. Do not rely on what the policy was last year.
Most universities updated their AI policies between 2023 and 2026. Many now have:
- A blanket policy in the student handbook or academic integrity office page.
- Course-level rules set by individual instructors (sometimes more permissive, sometimes stricter than the institution).
- Assignment-specific rules that override both.
Do this today, before your next assignment:
- Search your university's website for "AI policy" or "generative AI" or "academic integrity AI."
- Read the policy in full. Note any disclosure requirements.
- Check the syllabus of every course you are taking. Look for an "AI policy" section.
- If a course has no policy, email the instructor and ask. Save the reply.
You want a paper trail. If you ever face an integrity hearing, "I asked and was told it was allowed" is a strong defense; "I assumed" is not.
Disclosing Your AI Use
Many institutions and journals now require an "AI use statement" or methods disclosure. Even when not required, disclosing is the safer default — and it is rapidly becoming standard practice.
A simple disclosure might read:
"I used ChatGPT (GPT-5, OpenAI) to brainstorm initial topic angles and to summarize three of the papers I had already found through Semantic Scholar. All sources are real and were read in full. The argument and writing are my own."
Or, for a journal submission:
"Generative AI (Claude 4.7, Anthropic) was used to check grammar and suggest alternative phrasings. No content, references, or analysis was generated by AI."
A good disclosure says: which tool, which model version if known, for which tasks, and what you did NOT use it for. The clearer your boundaries, the more trust you build with your readers and reviewers.
The Five Red Lines
These are non-negotiable. Crossing any one of them can end your academic career.
1. Never submit AI-generated text as your own. Even if you edit it lightly. The originality of the analysis must be yours.
2. Never include a citation you have not personally verified. AI tools fabricate references. Every author, year, title, and journal must check out against a real database (Google Scholar, Semantic Scholar, your library catalog).
3. Never use AI to circumvent a clear assignment requirement. If the assignment is "summarize the readings in your own words to demonstrate comprehension", asking AI to summarize them defeats the entire purpose.
4. Never share your university credentials with AI tools, or paste in confidential research data. Some chat tools train on your inputs. If your research involves human subjects, proprietary data, or unpublished work from a collaborator, treat the chat box as public.
5. Never lie about your AI use. If you used it, disclose it. If you are caught hiding it, the penalty is almost always worse than disclosure would have been.
The "Defend It" Test
Here is a simple check before you submit any work.
Imagine your professor stops you after class and says: "Walk me through your argument. Why this thesis? Why these sources? What does Smith (2022) actually argue, and why does it support your point?"
If you can answer in detail, off the cuff, your AI use was fine. If you would have to look back at the paper to remember what is in it, something has gone wrong.
This is sometimes called the "defend it" test, and it has been used to catch AI misconduct in vivas, oral exams, and class discussions for years now. Many professors are starting to do random oral spot-checks for exactly this reason.
A Practical Exercise
For your next assignment, before you start, write a one-paragraph "AI use plan" in a text file:
- "I plan to use [tool] for [specific tasks]"
- "I plan to NOT use AI for [specific tasks]"
- "I will verify every citation by [method]"
- "I will disclose my use by [method]"
Keep this file. If you ever need to defend your work, this plan plus the chat logs are gold.
Key Takeaways
- Use AI as a tool to support your thinking, not as a ghostwriter to replace it.
- Find your university's AI policy and your course's specific rules — in writing — before each assignment.
- Disclose your AI use clearly and conservatively. When in doubt, disclose more, not less.
- Never cross the five red lines: do not submit AI text, do not cite unverified sources, do not circumvent assignment goals, do not share confidential data, and do not lie about your use.
- Use the "defend it" test before submission: if you cannot explain your own work in detail, something has gone wrong.