Academic risks in the use of artificial intelligence (AI) by students of public universities

Abstract

The objective was to identify the academic risks associated with the use of Artificial Intelligence (AI) for school activities from the perspective of university students.  It was based on UNESCO´s policy of human-centered pedagogical approaches and learning improvement. The study employed a qualitative-phenomenological approach, with 58 students from 16 public univesities in Colombia, Brazil, Mexico, Argentina, Chile and Italy participating. Participants met the inclusion criteria: being enrolled in courses and having used AI in the previous year. A projective construction technique was used to collect data through a topic guide that elicited narratives. This guide was incorporated into Google Forms and distributed via WhatsApp. A content analysis of the narratives was performed using MAXQDA2024, assisted by visual tools such as MAXMaps and code frequencies. Fifty-seven coded segments were obtained, revealing three categories of academic risks, ordered from highest to lowest saturation: limitation of learning by AI, referring to the loss of reasoning skills; plagiarism by AI, including unreliable sources and plagiarism; and dependence on AI, referring to the perceived need for its use. The study concludes that the students' narratives reflect concerns related to cognitive development, academic ethics, and intellectual autonomy. This research provides evidence for the design of regulated pedagogical strategies and educational policies aligned with the international movement focused on maximizing the benefits and mitigating the risks of AI use.