Physician and the Machine: An Analysis of the Ethical Educational Use of Artificial Intelligence by Medical Students

BackgroundArtificial intelligence (AI) has been increasingly integrated into medical education, particularly through tools such as ChatGPT (OpenAI, California, US), Gemini (Google DeepMind, California, US), DeepSeek (Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., Hangzhou, China), and Microsoft Copilot (Microsoft Corp., Redmond, WA, US).Although these resources present the potential to expand access to information and assist in academic development, their use also raises ethical concerns related to the reliability of information, technological dependence, plagiarism, superficial learning, and potential impairment of clinical reasoning.This study aimed to investigate the perception of medical students regarding the use of AI in their academic training, emphasizing perceived ethical limits and practices considered ideal for its utilization. MethodsThis was a quantitative, descriptive, and cross-sectional study conducted with medical students using a structured questionnaire administered via a digital form. ResultsThe sample consisted of 120 participants, of whom 99.2% (n=119) reported having previously used AI tools for academic purposes.It was observed that 73.3% (n=88) of the students use AI recurrently, primarily for summarizing content, researching medical information, simulating clinical cases, resolving questions, drafting academic texts, and planning studies.Most participants acknowledged the positive contribution of AI to learning, with 60.0% (n=72) stating that these tools contribute significantly and 36.7% (n=44) stating they contribute to a limited extent.However, students also identified relevant risks, especially the reduction of their own clinical reasoning, technological dependence, lack of information reliability, superficial learning, and academic plagiarism.Furthermore, 83.3% (n=100) stated that they did not receive adequate institutional guidance on the use of these tools, and 67.5% (n=81) demonstrated an interest in receiving further training on the ethical and pedagogical use of AI.No participant (0.0%, n=0) advocated for the prohibition of AI in academic contexts, while 67.5% (n=81) considered that its use should be regulated with ethical and pedagogical criteria. ConclusionsAI already significantly integrates the academic routine of medical students and is perceived as a useful resource for learning, but it is still used in a scenario of insufficient institutional guidance.The findings reinforce the need for curricular guidelines, faculty training, and specific ethical education to promote the critical, responsible, and pedagogically appropriate use of AI in medical training.