Investigating Lived Experiences of Trusting AI to Misconceptions in Karachi, Pakistan, at the University level: An Interpretative Analysis

Abstract

In this study, university students in Karachi are implicitly analyzed in terms of beliefs in, information interpretation, and response to artificial-intelligence-generated fake content in the academic setting. The purposive sampling made possible the stimulated-recall type of interviews based on the interpretative phenomenological analysis (IPA) when each participant read one or two academic responses provided by AI and then thought over them in semi-structured Zoom interview. Thematic analysis, in keeping with the Brain and Clark paradigm, results in students considering AI to be a convenient scholarly resource and they trust the results of the latter due to their disguised character hence breeding the formation of misconceptions at the unconscious level. The cases of error detection occurred when the responses of AI were opposed to the existing information. These results indicate the need to introduce AI-literacy courses in universities and set ethical standards to students. This study shows the necessity of empirical testing in the research of the future with the use of interventions related to literacy by referring to the example of Karachi.