May Tech Tips: AI Detection through Turnitin

As the use of generative AI rises, instructors are now facing a common issue – how to detect when a student may have potentially used AI in an assignment. ChatGPT has become a popular text generative AI that can generate responses based on a user’s prompt. Turnitin has introduced an AI writing indicator to its toolkit. Now when Turnitin has been enabled for a file submission assignment, the similarity report will include an AI writing detection section. Clicking on the tab will show a preview of the document with text highlighted if it's detected to be potentially AI-generated. 

The similarity score and the AI writing detection model are separate from each other. The AI writing highlights are not visible in the Similarity Report. To view the text highlighted by the AI detection model, the instructor must click on the AI writing indicator. 

Turnitin’s AI writing detection model only highlights text that is highly likely to be AI-generated. This is to help with ensuring that students are treated fairly while also safeguarding the institution's academic integrity standards. When using the AI detection tool, keep in mind that the percentage is interpretive and should not be used as a definitive measure of academic misconduct, as there might be a possibility of false positives. Nonetheless, it can be used to help instructors decide how to best handle student submissions that may have been produced, even in part, by AI writing tools. 

The Office of Digital Learning additionally has a knowledge article available for instructors to refer to. For more information about Turnitin’s AI writing detection capabilities, please see Turntin’s feature page.