"As you may know already, Turnitin recently added an AI detector to their service. This came as a surprise to both me and our Turnitin administrator as we were not expecting this to be available until early next year. This early release has many individuals concerned across the country.
I have concerns about the reliability of the AI detector, especially because it cannot be verified by the professor or myself. Normally, we can verify the suspected plagiarized portion of a paper by consulting the online sources which Turnitin provides. However, with AI writing, the only “evidence” will be a percentage score—a score that is generated from another AI program. Even Turnitin states that those percentage scores should be treated as indicators and are not definitive (https://www.turnitin.com/
Since the reliability of the AI detector is in question, at this time my office will not be accepting cases based solely on the AI detector report. I recommend that faculty members speak to students about their expectations concerning the use of AI. If you receive a high AI score from Turnitin, one option would be to have the student rewrite the assignment. Of course, grades are always determined by faculty.
New technology is being developed as we speak, and AI detectors will become more reliable over time. I will keep discussing this with Academic Integrity colleagues across the country, ITD, and Turnitin and will keep you updated as best as I can. I look forward to future discussions with faculty about AI and AI detectors in the coming academic year."
No comments:
Post a Comment