Use and perception of artificial intelligence among Spanish medical students.

Abstract

Introduction: The recent emergence of artificial intelligence (AI) tools, especially generative language models, is transforming university environments and posing pedagogical, ethical, and professional challenges in medical education. However, few studies analyze students' experiences and perceptions in the Spanish context. Objective: To explore the use and perception of AI among medical students in Spanish medical schools. Methods: A multicenter cross-sectional study was conducted using an anonymous online questionnaire administered to students in Spanish medical schools (February–March 2024). Descriptive statistics of the collected variables and inferential analyses were performed. Open-ended responses were subjected to exploratory inductive thematic analysis to identify recurring patterns of meaning. Results: A total of 508 students from 32 medical schools responded to the survey (67.5% women). 73% reported using AI-based applications in their daily lives and 68% in their academic activities. The most frequent uses were comparing information (67.7%), preparing for exams (53.1%), and proposing solutions to tasks (51.4%). ChatGPT was the most used tool (80.8%). Although adoption was high, the perceived impact on daily life was moderate (mean 2.39/5). More than 70% expressed concern about the accuracy and quality of the information generated. 81% had not received specific training, and 86% expressed a desire to receive it. The qualitative analysis identified five main themes: the need for regulation, concern about biases, the perception of AI as a complementary tool, the demand for structured training, and the assessment of its usefulness as explanatory support. Conclusions: Spanish medical students show a broad but critical adoption of AI, combining frequent use with an awareness of its limitations. The high demand for specific training suggests the need to integrate competencies related to the responsible use of AI into medical curricula.