Earlier this term, I attended a webinar featuring Notre Dame’s James Lang (James Lang’s Profile Page). The title of the presentation was, “Academic Integrity in the Age of AI.” At one point during the presentation, he had received so many questions about how to prevent students from using Generative AI on this or that assignment that he responded to the moderator’s summarizing question in a memorable way: “If you came here looking for a strategy to foil AI, you’re not going to find it here. You’re not going to find it anywhere.” 

 

It was a pretty interesting moment that highlighted some of the central challenges and anxieties Generative AI is creating in higher education. 

 

This spring marks the first full year of AI on campus and, so, it felt appropriate to share a little about where the issue stands currently and what you can do about encouraging authentic engagement with learning in your courses. 

 

First, it’s important to point out that the landscape has shifted significantly. Other products have appeared and matured since ChatGPT launched in late 2022/early 2023. 

• Open AI’s paid version, GPT 4, still looms large in the space. 
• Microsoft’s Bing with GPT 4 has been renamed “Microsoft Co-Pilot” and is being integrated into its Office Suite of programs. 
• Google’s Bard has been upgraded and renamed “Gemini”

 

These have all become more powerful and capable. 

 

A recent post by Ethan Mollick illustrated how much Generative AI can accomplish in a mere 59 seconds (Mollick Article). It seems clear that these technologies will continue to disrupt and affect higher education, the workplace, and how we communicate for the foreseeable future. 

 

Detection Technologies Prove to be Inadequate

Last year, the common belief was that the concern over AI and academic integrity would be short-lived once detection software came to the market. Something else happened instead. Those technologies failed to produce reliable results OR gave results with little insight into their methodology and inner-workings. It appears that easy detection of AI generated text will remain elusive for the foreseeable future. 

 

What is Being Done/ Can be Done?

• CSU has responded with a series of Provost’s Colloquia on the issue. Provost’s Colloquium Information and Videos
• Faculty-led groups have been formed across campus, including the TILT Community of Practice on Generative AI (Community of Practice page for TILT) and the College of Liberal Arts Generative AI Task Force, who recently created a valuable memorandum on the topic.  
• Pedagogical approaches to co-existing with Generative AI are the hot topic of the professional development- Higher Ed landscape. I encourage everyone to consider upcoming webinars offered by organizations such as the International Center for Academic Integrity and others such as this recent one sponsored by AAC&U (Recording of AAC&U Webinar). 

 

Last Thoughts

Going back to that webinar I mentioned earlier, Dr. Lang referenced a 2023 study (Impact of Several Low-Effort Cheating-Reduction Methods in a CS1 Class) that showed some evidence for the efficacy of so-called “low effort cheating reduction methods.” These methods are similar to those long recommended by the CSU Academic Integrity Program and include: 

• Talking with your class about academic integrity
• Having students complete an integrity quiz or tutorial. 
• Reminding students about your expectations throughout the semester.
• Making students aware of all the help resources available on campus.