If your work is flagged by an AI detector

What is known about detectors

• A Stanford study found that GPT detectors misclassified writing by non-native English speakers as AI-generated far more often than writing by native speakers (Liang et al., 2023).
• OpenAI withdrew its own AI text classifier in 2023, citing its low rate of accuracy.
• University guidance, such as Michigan State University's, says detection tools should not be used as the sole basis for adverse action against a student.

What you can do

1. Stay calm and read your institution's process. Find the academic integrity policy and your rights in it. Many campuses have an ombuds office or student advocate who can advise you.
2. Gather your process evidence. Drafts, version history in Google Docs or Word, notes, outlines, code commits, lab notebooks, and sources you read. These show how the work developed over time.
3. Bring your Usewright record if you have one. Your project record and AI judgment record show, with dates, what you planned and what you did. They are your own account recorded at the time, so treat them as supporting context, not proof. You decide whether to share them.
4. Offer to explain your work. Being able to talk through your reasoning without notes is often the most convincing evidence. Practice with the questions on your project record page.
5. If you did use AI, say so plainly and explain how. A clear, honest account is better for you than a denial that falls apart.

A message you can send

Hello, I understand my submission was flagged by an AI detection tool. I would like to understand the concern and show how I did the work. I can share my drafts and version history, my notes, and a dated record of how I used or did not use AI. I am also glad to explain my work in person. Could we find a time to talk? Thank you.