Framing the threat posed by ChatGPT for online assessments: Incorporating Discord and deep learning

Researchers have found that GenAI places considerable pressure on the integrity of online summative assessments. My research strengthens this conclusion in two ways. I argue that the impact of GenAI for online assessments is as widespread as are already popular messaging applications such as Discord. I argue furthermore, on the basis of a study utilizing problem sets from my own PHIL 007 Introduction to Critical Thinking, that utilization of GenAI on online summative assessments is not necessarily mitigated by devising problem sets which test for higher-order learning. ChatGPT-4o performed swimmingly on problem sets that consisted almost entirely of challenging critical thinking exercises. While these two results may seem to entail a dire situation for online summative assessments, the way forward for higher education in online spaces will require reshaping assessment validity so that learning outcomes are not compromised by student usage of GenAI.