On April 4th, Turnitin released an artificial intelligence writing detection feature intended to identify text generated by chatbot/text generators such as ChatGPT.
This feature is not currently available to Berkeley instructors as campus leadership has opted out of this feature until we have the opportunity to review the feature internally and vet it for privacy, security, and accessibility, as well as for general functionality.
UC Berkeley has opted out of this feature, along with several peer institutions, for a number of reasons:
-
Inability to test before launch
-
Instructional support staff did not have an opportunity to review the feature before it was released to all users.
-
Timing
-
The feature was released mid-semester, shortly before final exams.
-
Student access
-
Unlike the standard plagiarism detection features which can be made visible to students to assist with properly citing material, this feature is only available to instructors.
-
Unclear actionability
-
Turnitin’s standard plagiarism detection features enable instructors to point to a specific source. The AI detection provides no such conclusive proof, and it is unclear how these findings would or should be used in an academic integrity investigation.
Next steps
We are currently offering access to this feature via an opt-in pilot. Please contact academicintegrity@berkeley.edu if you would like to participate, or if you have any questions about this topic.
For more information, see this resource from the Center for Teaching and Learning: Understanding AI Writing Tools and their Uses for Teaching and Learning at UC Berkeley