Tertiary Education Integrity in a ChatGPT Conscious World: Preliminary Kenyan Observations

The world has gained a keen and ubiquitous interest in the role that artificial intelligence (AI) will play in future everyday life. Many previously niche tasks have seemingly and suddenly become machine realizable through the use of machine models. This paper presents a brief investigation of how one such novel model - ChatGPT - has been adopted by tertiary education students. The data was generated by applying a single essay-type question to 572 first-year undergraduate students at the University of Nairobi. From the results collected, it was observed that in the two months since the model's inception, over 30% of the student body selected was aware of its existence and actively used it in an assessment space. From the niche nature of the question, AI detection was found to be response length dependent. Some qualitative properties of AI-generated answers were also established.