When university students’ insecurities drive them to rely excessively on Artificial Intelligence: Relationship between fear of failure, self-esteem, social support and ChatGPT dependence

Abstract In learning processes, students with a high fear of failure and low self-esteem may be more inclined to use technologies like artificial intelligence (ChatGPT…) to avoid making mistakes, potentially even developing a dependency on such tools. However, it is believed that having peers and teachers who help these students build confidence and receive support to overcome failure could act as a buffer against attachment to technological tools like ChatGPT. To test this hypothesis, 467 university students participated by completing a series of scales to measure their fear of failure, self-esteem, peer support, teacher support, and dependency on ChatGPT. Following a moderated mediation analysis, the results revealed that the initial hypothesis was not supported. Indeed, higher levels of fear of failure and lower levels of self-esteem did not lead to an increase in ChatGPT dependence. Likewise, social support from peers and teachers did not influence the previously mentioned relationship.