AI Detection Added to Turnitin

Turnitin has announced the launch of its AI writing detection capabilities. This feature will help instructors identify when AI writing tools such as ChatGPT may have been used to write any part of the content submitted in a student’s assignment.

Beginning April 4, this capability will be temporarily available through December 2023 within the products we use at Baylor - namely Turnitin Feedback Studio. This capability is being activated by Turnitin now in order to provide an immediate response to the perceived rise in AI-generated assignments. Please keep in mind that the AI detection results will not be visible to students; only instructors and administrators will be able to see the results of the detection tool. Additionally, Turnitin will only be able to process submissions written in long-form English for AI detection.

AI detection has been added to the Turnitin Similarity report, which means that your existing workflow will remain unchanged and you can also access it via Canvas. In-product guidance and access to product help resources will be at hand to explain the feature. The tool provides a percentage, which indicates the amount of qualifying text within the submission that Turnitin’s AI writing detection model determines was generated by AI (with 98% confidence based on data that was carefully collected and verified in a controlled lab environment). This qualifying text includes only prose sentences, meaning that the tool only analyzes blocks of text that are written in standard grammatical sentences. The detection functionality does not include the recognition of other types of writing such as lists, bullet points, or other non-sentence structures.

To prepare you further for the use of this feature and to help you understand how you should interpret the results, Turnitin has put together the following guidance and resources.

AI Detectors like this one from Turnitin can be helpful when determining whether or not a conversation with a student is needed. In keeping with the University’s Honor Code, each case of an alleged Honor Code violation will be considered based on the facts of that individual case.

For additional suggestions and recommendations, please refer to the Office of Academic Integrity’s page Suggestions for Dealing with Artificial Intelligence (AI)-Produced Assignments.