Artificial intelligence (AI) is significantly transforming education, offering both innovative opportunities and significant challenges. This paper addresses this issue by describing an experience carried out with third-year students of the Bachelor's Degree in Business Administration and Management in a subject in the area of applied economics at the University of Murcia (Spain). The experience consisted of offering a task in which students had to answer questions about an article with the help of AI. The students should not simply copy the responses generated by the AI; rather, they had to obtain their own answers with the help of the AI and then answer questions about their experience. Three questionnaires were given to the students: one prior to the task to analyze their starting situation, one for the task, and one after the task to analyze the situation of the students after it.

The questionnaires before and after the task show, first, the existence of differences between men and women. Women show less confidence in AI, less interest in using it in the future and less use of AI than men do for academic matters but greater use for personal matters; however, both sexes make more use of AI for academic matters. Additionally, men are more likely to use AI in the future. Second, an increase in the assessment of the usefulness and reliability of AI is observed after performing the task, with the increase in utility greater in the case of women. Third, among the different learning options for the use of AI, students prefer to be trained rather than learning on their own, showing a greater preference for the use of AI in the activities of the subjects than having a specific subject devoted to the use of IA; the difference between these two options is less in the case of women.

In the task questionnaire, the students indicated that they had contrasted the AI responses mainly by comparing them with the article provided and, to a much lesser extent, by comparing the responses of different AIs and looking for information in other sources. Regarding the errors found, the students mentioned the lack of updating of the information provided by the AI, too general or schematic answers, and problems with the precision and quality of the information. In addition, some students indicated that AI did not use economic terms appropriately. Notably, students who used multiple sources to contrast AI responses tended to detect more errors. In relation to the usefulness of AI at the university, the students highlighted its use to have personalized tutorials, followed by the search for and analysis of information, the completion of tasks and the writing and correction of papers.

The results of this project reveal several questions of interest, such as the existence of differences in the assessment and use of AI between men and women, that, at least in the advanced courses of the degree, students are aware of its risks and limitations but that can be overcome through contrasting and that part of the initial mistrust is attenuated with the use of these tools. Their almost absolute conviction that AI is useful and their clear commitment to use them more in the future mean that this reality cannot be ignored by the educational community. In this sense, it is crucial to encourage the responsible use of AI, emphasizing the importance of critical thinking, creativity, originality and academic integrity.

Keywords: AI literacy, AI tools, student-centered learning, innovative teaching, Economics.