Skip to Main Content

Artificial Intelligence: ChatGPT and Beyond: AI: Plagiarism

Is the use of AI plagiarism?

When might the use of AI tools constitute plagiarism?  If you put in a prompt for an AI tool and turn in the results as your own, isn't that the same as presenting someone else's work as your own?  Isn't that plagiarism?   What about learning to use AI?  Is that a valid  reason to use AI to assist with a course assignment?  It depends!  Check with your instructor before using AI !   Also, if you use AI, be sure that you are willing to be accountable for the results!   Check for accuracy. 

Ditch That Textbook. How to define "cheating" and "plagiarism" with AI. 

Eaton, Lance.  Classroom policies for AI generative tools 

Scribbr.  University policies on AI writing tools: Overview and list.  

------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Lyda, M. (2022, December 15).  Plagiarizing Chat GPT - Is it illegal?  YouTube.  (auto-generated captions) [4:43 minutes] [Chat GPT wrote the entire script] 

Stern, J. (2022, December 21).  Cheating with ChatGPT: Can openAI’s chatbot pass AP lit?  Wall Street Journal.  YouTube [6.58 minutes] 

Test your abilities! 

Real or Fake Text? Play by Category.  How good are you (and other humans) at knowing when text has been written by a computer?   Try it out and then read the research paper about this test:

Dugan, Liam, et al. "Real or fake text?: Investigating human ability to detect boundaries between human-written and machine-generated text." Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 37. No. 11. 2023. https://ojs.aaai.org/index.php/AAAI/article/view/26501/26273

 

Tools to Catch AI Plagiarism

ANOTHER UPDATE: 

Open AI writes "Can I ask ChatGPT if it wrong something?" The  short answer is "no!" 

 

UPDATE.  

Bailey, Jonathan. (2023, July 27).  The current state of detecting AI writing.  PlagiarismToday.  https://www.plagiarismtoday.com/2023/07/27/the-current-state-of-detecting-ai-writing/  "There is functionally no reliable detection of AI works, and there’s no clear solution on the horizon."

A more trustworthy AI detector? Maybe!  “UC Berkeley Researchers Introduce Ghostbuster: A SOTA AI Method for Detecting LLM-Generated Text. Ghostbuster utilizes a three-stage process involving probability computation, feature selection, and classifier training. It significantly outperforms existing models in detecting AI-generated content, achieving a 97.0 F1 score across various conditions. This tool addresses concerns about the authenticity of AI-generated text, particularly in academic and journalistic contexts.” Reported in AI Breakfast  November 22, 2023

Example of tools set up to detect use of AI: ( "Exercise caution when using these tools to detect academic dishonesty."  OpenAI Text Classifier in 'limitations" notes that it is not fully reliable.) 

AI Writing Check (Quill.org and CommonLit.org) 

GLTI (from MIT_IBM Watson AI Lab and HarvardNLP) 

GPTZero: Created by Edward Tian. edward_the6 (@Edward Tian)  Try the beta at  https://t.co/220Y32rS4z or https://t.co/vbbM1KUC5X  (Tweet from January 2, 2023)  Updates: GPTZ0 substack.  GPTZero will include ability Word and PDF uploading; improved ability to handle text that combines AI and human writing together, February 20, 2023.  GPTZero is partnering with K16 Solutions to offer integration with Canvas.  

OpenAI: New AI classifier for indicating AI-written text 

Quillbot AI Plagiarism Checker  

Turnitin announces AI writing detector and AI writing resource center for educators.  February 13, 2023  Turnitin promises 97 percent accuracy with less than 1/100 false positive rate.  

 

 

Plagiarism and Cheating Instances

Can you outwit AI plagiarism detectors?

AI Advantage. (2023, January 8).   How to get around ChatGPT detectors.  YouTube.  [On use of ChatGPT and plagiarism detection. Can you ask an AI to rephrase generated text?  Yes.   Can you ask an AI to write in your own voice by submitting some of your own writing?  Yes.  Can you fool some of the current plagiarism detectors? Yes.  Will you be able to count on that in the future? No!] [Igor Pogany]

Gewitz, D. (2023, January 13).  Can AI detectors save us from ChatGPT?  I tried 3 online tools to find out.  ZDNET.

Hirst, T.  (2022, December 12).  Can ChatGPT detect GPT3 generated texts?  OUseful.Info. 

Nolan, B. (2023, January 14, 2023 Two professors who say they caught students cheating on essays with ChatGPT explain why AI plagiarism can be hard to prove.  Business Insider.

van Oijen, Vivian (2023, March 31).  AI-generated text detectors: Do they work?   Surf Communities. [Comparison of seven different free AI-generated text detectors]. 

For instructors

Phillip Dawson (@phillipdawson) on Twitter has some reminders to instructors about using plagiarism tools: 
Did the student give permission to upload their work?  
What are the legal, ethical obligations around student data and privacy?  Does the assignment include student names or sensitive information? 
Will student data be stored?  Can you or students ask that the data be deleted?  
What is the data about accuracy of the tool?  What if the results are not trustworthy? 
Do false positives cluster around certain demographics? 
From https://twitter.com/phillipdawson/status/1615474118345859082 
January 17, 2023

 

Cotton, Debby R. E., Cotton, Peter A., & Shipway, J. Reuben.  (2023, March 13).  Chatting and cheating: Ensuring academic integrity in the era of ChatGPT.  Innovations in Education and Teaching International. An article written in part with Chat-GPT.  Is that cheating?  

A Different Approach