Scholarlink

Background <p>As AI integration in medicine becomes critical, this study evaluated medical students’ attitudes and knowledge regarding AI and determined its impact on career specialization choices.</p> Method <p>This cross-sectional study used a voluntary, web-based convenience sample of 274 first- to fifth-year students at Gazi University Faculty of Medicine in Ankara, Türkiye (October–November 2025). Data were collected with a structured questionnaire incorporating the validated Perceptions on Artificial Intelligence in Medicine (PAIM) scale.</p> Results <p>Although 74.5% reported no formal AI training, 52.6% expected AI to influence their specialization choice. Students who expected AI to shape their careers reported higher “Knowledge and Trust” (<i>p =</i> 0.018); the “Disadvantages and Risks” subscale did not differ significantly between career-impact expectation groups (<i>p =</i> 0.131). Among the three PAIM dimensions, only “Informed Self-Control” differed significantly by training phase, with clinical-phase students scoring lower than pre-clinical peers (<i>p =</i> 0.027). In multivariable regression, expecting AI to influence specialization was the only positive predictor of “Knowledge and Trust” (B = 0.16, <i>p =</i> 0.018), and formal training predicted higher “Informed Self-Control” (B = 0.28, <i>p =</i> 0.015), while older age predicted lower scores (B = − 0.05, <i>p =</i> 0.015). Both models were statistically significant but explained only a small share of variance (adjusted R2 = 0.02 and 0.04). Overall, 94.2% supported integrating AI into the medical curriculum.</p> Conclusion <p>The findings highlight a marked gap between students’ strong demand for AI training and the limited formal instruction currently provided. Cross-sectional differences across training phases may reflect a shift from early optimism toward a more cautious stance, although the design cannot establish such a developmental trajectory. Because more than half of students expected AI to affect their specialty choice, curricula should move beyond technical instruction to integrate ethical and clinical reasoning alongside AI literacy. These associations are exploratory and require confirmation in larger, longitudinal studies.</p> Trial registration <p>This study did not involve a clinical trial and therefore was not registered in a clinical trials registry.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

From technological optimism to clinical realism: medical students’ attitudes toward artificial ıntelligence and ıts career ımplications — a cross-sectional study at a Turkish Medical Faculty

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月1日

查看全文
  • Mehmet Koca,
  • Sevilay Ulas

摘要

Background

As AI integration in medicine becomes critical, this study evaluated medical students’ attitudes and knowledge regarding AI and determined its impact on career specialization choices.

Method

This cross-sectional study used a voluntary, web-based convenience sample of 274 first- to fifth-year students at Gazi University Faculty of Medicine in Ankara, Türkiye (October–November 2025). Data were collected with a structured questionnaire incorporating the validated Perceptions on Artificial Intelligence in Medicine (PAIM) scale.

Results

Although 74.5% reported no formal AI training, 52.6% expected AI to influence their specialization choice. Students who expected AI to shape their careers reported higher “Knowledge and Trust” (p = 0.018); the “Disadvantages and Risks” subscale did not differ significantly between career-impact expectation groups (p = 0.131). Among the three PAIM dimensions, only “Informed Self-Control” differed significantly by training phase, with clinical-phase students scoring lower than pre-clinical peers (p = 0.027). In multivariable regression, expecting AI to influence specialization was the only positive predictor of “Knowledge and Trust” (B = 0.16, p = 0.018), and formal training predicted higher “Informed Self-Control” (B = 0.28, p = 0.015), while older age predicted lower scores (B = − 0.05, p = 0.015). Both models were statistically significant but explained only a small share of variance (adjusted R2 = 0.02 and 0.04). Overall, 94.2% supported integrating AI into the medical curriculum.

Conclusion

The findings highlight a marked gap between students’ strong demand for AI training and the limited formal instruction currently provided. Cross-sectional differences across training phases may reflect a shift from early optimism toward a more cautious stance, although the design cannot establish such a developmental trajectory. Because more than half of students expected AI to affect their specialty choice, curricula should move beyond technical instruction to integrate ethical and clinical reasoning alongside AI literacy. These associations are exploratory and require confirmation in larger, longitudinal studies.

Trial registration

This study did not involve a clinical trial and therefore was not registered in a clinical trials registry.