ChatGPT and Generative AIs: What It Means for Academic Assessments

Generative Artificial Intelligence (GAI) is a subfield of AI that involves generating machines or algorithms that allow creating new and original content. GAI is receiving huge popularity with Open AI making ChatGPT available to the public as of 30 Nov 2022. Higher education is one of the industries which is being impacted most by GAI, and there is huge uncertainty around what GAI mean for universities in terms of assessments and academic integrity. Given this situation, this article intended to explore what academic assessments are most vulnerable to GAI, which ones are least vulnerable, offered ranking among those assessments, and offered strategies as to what universities can do to ensure GAIs provide least disruption to their business and academic assessment integrity. Data was collected by interviewing 4 GAIs namely ChatGPT, YouChat, Chatosonic AI, and Perplexity AI. The interview included asking multiple open-ended questions. The results found the following assessments as the most vulnerable assessments against GAI, multiple-choice tests, short answer questions, True/False questions, essays, research papers and presentations. The findings also showed that the following assessments are the least vulnerable assessments against GAI, exams, practical assessments, performance-based assessments, open-ended questions, and group projects. These assessments were also ranked by the 4 interviewees, and recommendations for universities is provided.