Will faculty members be able to use software to identify auto-generated text from human writing?
Thus far, not reliably. Existing detectors such as OpenAI’s classifier and GPTZero have been shown to identify human-written text as “likely AI” a significant portion of the time.
False positives can lead to false accusations.
What other steps can be taken?
- Use Google Scholar to verify citations: At this time, many AIs are still creating fake citations. Double-checking citations from a paper by pasting the whole citation (or at least the article title) Google Scholar will verify if the citation is an actual resource, the majority of the time.
- Require students to submit the PDF of articles cited in their papers
- Create an assignment around an issue that is specific to the local community
- Direct students to write about very recent news events
- Have students show or explain their work
- Ask students to give an oral presentation, along with the written work
- Handwritten essays (owch!)
- Put project-based learning to work