From the Provost's Office - It is vital that faculty be very specific about when and how AI can be used. Otherwise, charges of academic dishonesty will be hard to uphold if appealed.
Currently, there is no equivalent of Turnitin for checking student work for the use of AI. There are beta versions of some tools like GPTZero and the ZeroGPT (Bowman, 2023) that purport to detect and report the likelihood that text has been generated using AI, however, these tools are not completely reliable and should not be used as plagiarism detectors. OpenAI, the company that created ChatGPT, is developing a prototype of a watermarking tool that will help flag text as AI-generated (Morrison, 2023). Turnitin is also in the process of enhancing its products’ AI-detection capabilities (Caren, 2022).
Many inclusive teaching strategies also encourage academic integrity:
Over the course of the last academic year, Penn College has experienced multiple cases of academic dishonesty related to unauthorized use of generative AI. While most cases relate specifically to submission of writing assignments generated using AI technologies, some have been used in the development of images and other more complex assignments.
Many AI “checkers” appear to be unreliable and may indicate a use of AI where there is none. Consider using multiple options to determine if a consistent finding is present (Turnitin, Sapling, ZeroGPT, GPTZero, etch). In addition to use of these tools, consider some of the following when attempting to determine if work is original: