AI detection with Turnitin 

As of Dec. 7, 2023

Update from the Normandale Campus Academic Technology Team (NCATT) and the Dean of Academic Services:

Note: the following email was sent on Dec. 7th, 2023, to faculty who currently use Turnitin-D2L integration.

Last April, Turnitin added an "AI Score" as a free trial of an upgrade that campuses can purchase. This trial expires at the end of December. Our NCATT (Normandale Campus Academic Technology Team) did a thorough review of this add-on product and has decided not to buy it. The team’s primary concern is about its level of accuracy. The tool has performed poorly when some of our folks tested it by generating content using different AI tools, and our experience is not unique: see this Washington Post article. This raises serious concerns about using a student’s AI score to initiate conversations with them about their writing process.  

If you have used this feature, please know that we will continue to monitor Turnitin's developing capabilities vis a vis generative AI and are willing to acquire reliable tools that can equitably help our students learn. The NCATT is devoted to supporting our students and employees with AI resources. Please go to the AI Avenue webpage for up-to-date information and resources.​​

• If you encounter a student's work that is questionable, it is highly recommended that you start a conversation with the student about their writing based on your assignment criteria and course objectives, without direct accusations of AI misuse.
• Consult the Dean of Students office if you would like assistance with a student integrity issue.

Overview for Instructors:

• For more info, read this Article on inequities of AI detection: 
  • “Programs to detect AI discriminate against non-native English speakers, shows study” - reporting by The Guardian on 10 July 2023, and as an open access journal article. 
  • The researchers “traced the discrimination to the way the detectors assess what is human and what is AI-generated.”
• The Normandale "AI Avenue" team will continue to provide resources for talking to students about AI, and/or adjusting assignments to generate authentic assessments.

What can Turnitin do that is reliable?

Turnitin's "Originality" checker is reliable as far as being able to cross-check references to online sources as well as Turnitin's "paper repository."

If you don't already have Turnitin enabled in D2L Assignments, download the PDF document in the "Attachments" section of this page for instructions.

Resources

1. Coldewey, D. (2023, July 25). OpenAI scuttles AI-written text detector over ‘low rate of accuracy’. Retrieved July 27, 2023, from https://techcrunch.com/2023/07/25/openai-scuttles-ai-written-text-detector-over-low-rate-of-accuracy 
2. De Vynck, G. (2023, May 30). ChatGPT ‘hallucinates.’ Some researchers worry it isn’t fixable. The Washington Post. Retrieved June 1, 2023, from https://www.washingtonpost.com/technology/2023/05/30/ai-chatbots-chatgpt-bard-trustworthy 
3. Edwards, B. (2023, July 14). Why AI detectors think the US Constitution was written by AI. Retrieved July 28, 2023, from https://arstechnica.com/information-technology/2023/07/why-ai-detectors-think-the-us-constitution-was-written-by-ai/3 
4. Fowler, G. A. (2023, April 3). We tested a new ChatGPT-detector for teachers. It flagged an innocent student. Retrieved April 4, 2023, from https://www.washingtonpost.com/technology/2023/04/01/chatgpt-cheating-detection-turnitin 
5. Fowler, G. A. (2023, Aug. 14) What to do when you’re Accused of AI Cheating: AI Detectors Like Turnitin and GPTZero Suffer from False Positives that can Accuse Innocent Students of Cheating. Here’s the Advice of Academics, AI Scientists and Students on how to Deal with it. ProQuest, https://ndcproxy.mnpals.net/login?url=https://www.proquest.com/blogs-podcasts-websites/what-do-when-you-re-accused-ai-cheating/docview/2850411186/se-2. 
6. Klee, M. (2023, June 6). She Was Falsely Accused of Cheating With AI — And She Won’t Be the Last. Retrieved June 7, 2023, from https://www.rollingstone.com/culture/culture-features/student-accused-ai-cheating-turnitin-1234747351
7. Myers, A. (2023, May 15). AI-Detectors Biased Against Non-Native English Writers. Stanford University Institute for Human-Centered Artificial Intelligence. Retrieved July 18, 2023, from https://hai.stanford.edu/news/ai-detectors-biased-against-non-native-english-writers 
8. Sands, L. (2023, April 6). ChatGPT falsely told voters their mayor was jailed for bribery. He may sue. Retrieved May 30, 2023, from https://www.washingtonpost.com/technology/2023/04/06/chatgpt-australia-mayor-lawsuit-lies