The Academic Use of Artificial Intelligence Tools Among First-Year Medical Students: A Cross-Sectional Online Survey

Technical Summary: The Academic Use of Artificial Intelligence Tools Among First-Year Medical Students

Problem Statement
While artificial intelligence (AI) tools are rapidly reshaping global educational practices, there is a paucity of evidence regarding how first-year medical students specifically utilize these tools for academic purposes, particularly within the context of Türkiye. Although generative AI offers potential benefits such as simplifying complex concepts and facilitating self-directed learning, its integration raises significant concerns regarding reliability (e.g., "hallucinations" or fabricated references), academic integrity (plagiarism, authorship), and the ethical implications of overreliance. Current literature suggests that while medical students generally hold positive attitudes toward AI, structured education on AI literacy, critical appraisal, and responsible use remains limited in Turkish medical curricula. This study addresses the gap in understanding the specific usage patterns, frequencies, and perceptions of first-year medical students to inform future curriculum planning.

Methodology
The study employed a descriptive, cross-sectional online survey design conducted at the Faculty of Medicine of Recep Tayyip Erdoğan University.

• Participants: The study included 78 first-year medical students (aged 17–22) who voluntarily agreed to participate. The demographic breakdown was 52.6% male and 47.4% female.
• Data Collection: Researchers utilized a questionnaire developed after a literature review. The instrument assessed sociodemographic characteristics, daily internet usage duration, specific AI tools used, frequency and purposes of use, access devices, and student attitudes toward AI in academic learning.
• Analysis: Data were analyzed using IBM SPSS Statistics. As a descriptive study, results were summarized using frequencies, percentages, means, and standard deviations. No inferential statistical comparisons were performed.
• Ethical Considerations: The protocol was approved by the university's Health Sciences Ethics Committee (Approval No. 2026/108), and electronic informed consent was obtained from all participants in accordance with the Declaration of Helsinki.

Key Results

• Adoption and Tools: 100% of participants reported using AI tools. ChatGPT was the most frequently used application (62.8%), followed by Google Gemini (35.9%).
• Frequency: Usage was frequent, with 44.9% of students using AI tools several times per week and 30.8% using them daily.
• Primary Purposes: The dominant use case was "studying course content" (61.5%). Other purposes included solving practice questions (12.8%), understanding course topics (11.5%), and obtaining academic information (7.7%). Only 3.8% reported using AI for completing assignments.
• Access Devices: Smartphones were the primary access device (69.2%), followed by tablets (20.5%) and multiple devices (10.3%).
• Attitudes and Perceptions:
  • Students held generally positive views regarding the utility of AI. The highest agreement scores were for "AI tools save me time" (Mean: 4.18 ± 0.91) and "AI tools are helpful in understanding difficult subjects" (Mean: 4.13 ± 0.76).
  • Participants strongly agreed that AI-generated information must be verified (Mean: 3.83 ± 0.87) and that formal training on AI use should be provided (Mean: 3.69 ± 0.97).
  • There was broad consensus that AI tools should have a place in medical education (Mean: 3.85 ± 0.97).

Significance and Claims
The paper posits that AI tools have become an integral part of the academic routines of first-year medical students in Türkiye, primarily serving as support for self-directed learning and content comprehension. The study claims that while students perceive these tools as effective for saving time and clarifying difficult subjects, their awareness of the necessity to verify AI-generated content indicates a nascent but critical understanding of the technology's limitations.

The authors argue that these findings underscore the necessity of integrating AI literacy, critical appraisal, and responsible use into undergraduate medical education. Rather than viewing AI merely as a technical tool, the paper suggests it is an emerging competency area where future physicians must learn to evaluate outputs, recognize ethical implications, and distinguish between reliable and fabricated information. The study concludes that structured educational programs are required to transition students from passive users to critically engaged practitioners who can safely and ethically utilize AI in both academic and future clinical settings.

Limitations
The authors acknowledge that the study's generalizability is limited by its single-institution design and focus on first-year students only. Additionally, the reliance on self-reported data introduces potential response and recall biases, and the cross-sectional nature of the study precludes the establishment of causal relationships between AI use and academic performance.