Can We Detect AI?

The Good News: Sometimes

The Bad News: It is difficult to prove

Based on Palmer (2023):

If you use a chatbot detection tool, a fake rating is not enough evidence to accuse a student of cheating or plagiarism
Chatbot detection tools may violate GDPR privacy policies.
You will need to scrub any personally identifiable information.

Consibear
by F. Reichert using MidJoruney

Some student get caught ...

due to fake references

Source: Khrono - https://www.khrono.no/tre-studenter-ved-uib-ble-utestengt-etter-mistanke-om-ki-juks/828177

How to cover your tracks

Even worse News: There are tools to hide your tracks (see (Gluska 2023))

Chatbots are popular - 25% use them weekly

source: Khrono.no - Nov. 30, 2023 - https://www.khrono.no/hver-fjerde-student-bruker-chatgpt-ukentlig/829641

Source: NSO (Norsk studentorganisasjon) via Khrono

Conclusion: We have to change

Source: Miller (2023)

References

Palmer, K. (2023). Chatbot-Proof Assignments. Butler University Libraries. https://libguides.butler.edu/ai-in-the-classroom/chatbot-proof-assignments
Gluska, J. (2023, September 27). How to Bypass Turnitin AI Writing Detection. GoldPenguin Blog. https://goldpenguin.org/blog/how-to-bypass-turnitin-ai-writing-detection
Miller, M. (n.d.). Ditch That Textbook. https://ditchthattextbook.com
Miller, M. (2023, November 27). 6 ways to prepare students for an AI future. Ditch That Textbook. https://ditchthattextbook.com/ai-skills

Acknowledgments and Copyright

This Workshop/Course is sponsored by UiA PULS and the NORAD Tells Project 2021-2026