As AI literacy becomes an increasingly fundamental capability for future health professionals, it is essential that medical students are ready to navigate technological advances in their future workplaces. Evaluating the current level of AI literacy among medical students to support this goal is a necessary first step in designing targeted and effective educational interventions. In line with this objective, the study investigated AI literacy among medical students and investigated the association between attitudes towards AI, intention to use AI in clinical contexts, and individual characteristics such as previous AI training and AI interests.
To address this objective, snails were first verified through confirmation factor analysis in the Korean context. During this process, seven items were excluded and there was a final version consisting of 24 items (sanail-kr) that was used for analysis. Results revealed that students showed slightly below average levels of AI literacy. These findings are consistent with those of Laupichler et al. [6]reported comparable levels of AI literacy among German medical students, with a total average score of 3.76 (3.79 in this study). In their study, key ratings showed the highest mean score (M = 4.89, M = 4.64 in this study) and the lowest technical understanding (M = 2.63, M = 3.19 in this study), showing the same pattern as observed in this study. The relatively high scores on key ratings may reflect a further increase in media coverage on issues such as data privacy, ethics, and AI-related risks. However, given that the highest subscale scores only slightly above the midpoint of 4, the findings collectively highlight the need for more comprehensive AI literacy education. This interpretation is further supported by the fact that 85.17% of participants reported receiving less than 30 hours of AI-related training. Additionally, a systematic review by Kimiafar et al. [30] and Musabi Baigi et al. [2] Similarly, healthcare professionals and students generally expressed motivation to adopt AI, but they reported that they were inadequately trained and tended to have limited knowledge and skills when using AI. These findings highlight the urgent need to strengthen AI literacy education to enhance the findings of this study and better prepare future health professionals.
Furthermore, participants generally expressed positive attitudes towards AI. However, the average score for negative attitudes (reverse scores) was also slightly above the neutral midpoint. This suggests a balance between advantageous and careful perspectives. Students recognized the potential benefits of AI while acknowledging current limitations. This finding is consistent with the SEO and AHN survey results. [23]using the same scale to examine the attitudes of 230 Korean nursing students. On a 5-point Likert scale, the average positive attitude was 3.69, while the average negative attitude was 3.07. This reflects a similar pattern to that observed in this study. Similarly, Gordon et al. [1] A scoping review of attitudes towards AI application in clinical medicine reported the coexistence of support and concerns about AI among medical learners. Such concerns can affect learners who adopt a cautious attitude towards AI applications. Collectively, these findings highlight the importance of directly addressing both AI-related opportunities and concerns in future education and training programs.
Furthermore, the findings show that students with a higher level of AI literacy are more likely to maintain a positive attitude towards AI. Among the subcategories, practical applications show the strongest correlation with positive attitudes, suggesting that students who view AI favorably tend to be involved in it in their daily lives. Critical ratings were positively associated with positive attitudes and negative attitudes, and there was a significant correlation to positive attitudes. This means that the more students understand both the benefits and potential concerns of AI, the more likely they are to adopt a positive perspective. Furthermore, a positive association was observed between positive attitudes towards AI and previous AI training. Collectively, these results highlight the importance of structured AI literacy education that addresses both the benefits and potential risks associated with AI.
Regarding intentions to use AI in clinical contexts, participants reported scores slightly above the neutral midpoint. This intent was significantly associated with both AI literacy and positive attitudes towards AI. These findings show that AI literacy is an important factor affecting the intention of students to use AI in clinical practice, while also confirming previous studies showing a positive association between positive attitudes towards AI and intention to use it. [1, 13].
Furthermore, both AI literacy and positive attitudes towards AI were significantly correlated with student interests in AI and previous AI training, with a stronger correlation to AI interest than previous training. Interest in AI was also significantly related to intention to use AI, but previous AI training showed no significant correlation with this variable. Furthermore, within the AI literacy subcategory, practical applications showed a particularly strong correlation with student interest in AI than previous AI training. This suggests that students with a high interest in AI are more likely to engage in AI technology in their daily lives, thereby enhancing AI literacy and promoting positive attitude AI. The influential role of interest in AI in AI literacy is Laupichler et al. [6]reported that student interest in AI is more strongly associated with AI literacy than previous training experiences. Furthermore, there were significant differences in AI literacy between students with little training and students with less than 30 hours of training or more than 30 hours of training and students with mean differences of 0.68 and 1.1 points, respectively. These findings highlight the potential impact of AI education and suggest that programs offering over 30 hours of training may be necessary to meaningfully enhance students' AI literacy. The lack of a significant association between previous AI training and intentions to use AI may be attributed to small sample sizes, but further research is needed to clarify this relationship. Overall, the findings suggest that fostering AI interest is particularly effective in promoting AI literacy, and that educational interventions should aim to stimulate AI engagement and curiosity, particularly through programs that provide over 30 hours of training.
No significant differences were observed for any of the variables measured in this study by gender, year of study, or measure. These findings differ from those reported by Laupichler et al. [6]male students scored high scores in AI literacy and found that semesters were significantly associated with AI literacy levels. Similarly, Hashish and Arnajar [5]a study evaluating digital health literacy among nursing students, defined as perceived competence when using digital health tools, shows that senior students demonstrated higher literacy levels. Furthermore, Kwak et al. [13] Senior students reported significantly higher positive attitudes towards AI, and Sumengen et al. [20] Both found that male nursing students had a more positive attitude towards AI, as measured by the GAAI.
This discrepancy between these previous findings and current findings cannot be fully explained in this study. However, one possible explanation for the lack of significant gender differences is the voluntary participation in this survey and low response rates. Therefore, students who chose to participate regardless of gender may have had a greater interest in AI than the general student population. Furthermore, the lack of significant differences between study years may reflect limited integration of AI-related content into the curriculum. As mentioned before, Korean AI education is usually offered as an elective course only at the pre-medical stage. This may contribute to the lack of AI literacy and grade attitude variation. Alternatively, these insignificant results may be attributed to relatively small sample sizes. Regarding the lack of major differences, further research is needed to investigate whether and how AI literacy differs and changes depending on the discipline within healthcare education. To review and extend these findings, future research using larger and more diverse samples across multiple disciplines in medical education is needed.
The findings of this study highlight the urgent need to systematically integrate AI literacy into the health care curriculum. Currently, training opportunities are inadequate, with most students minimising exposure to AI education. Given the observed AI literacy among students over 30 hours of AI training, this study highlights the value of a sufficiently long structured program. Training over 30 hours is particularly effective in enhancing both AI literacy and positive attitudes towards AI. Furthermore, integrating AI content into the core curriculum can promote a more equitable and inclusive learning experience than offering it as an elective subject alone. Educators should also leverage students' existing interest in AI as a foundation for enhancing AI literacy. Educational programs must be purposefully designed to induce and maintain student interest in AI by incorporating practical activities using AI tools and by applying AI to solve real-world healthcare problems. Furthermore, given that students can approach AI with a cautious attitude towards AI applications, AI literacy programs must explicitly address concerns. Promoting open dialogue, promoting critical reflection, and exposure to practical applications can help learners develop a balanced, informed perspective on the role of AI in clinical practice.
Despite its contribution, this study has some limitations. First, the sample size was relatively small. The response rate for the online survey was very low due to two years of collective leave of Korean medical students caused by political issues. Furthermore, the sample was drawn from a single institution in one geographical area, limiting the generalizability of the findings. However, while participants may not fully represent all Korean medical students, Korean medical students tend to form relatively homogenous groups due to a highly standardized and competitive admission process. Furthermore, given the similarities in the curriculum of Korean medical schools, significant fluctuations in AI literacy between systems may be limited. Future research using larger and more diverse samples from multiple institutions is essential to verify and expand these results. Nevertheless, the present study provides preliminary insights that could serve as a valuable baseline for further investigation of AI literacy and attitudes among medical students. Furthermore, multinational studies can provide a useful comparative perspective on how AI literacy differs between educational and cultural contexts. Longitudinal or intervention-based studies examining the development of AI literacy over time may also inform future curriculum design.
Second, this study relies on self-reported data, which could be subject to response bias and subjective interpretation. To increase the validity of future findings, objective measures such as knowledge-based assessments and assessments of practical AI-related skills should be incorporated along with self-report devices. Finally, the adapted snail KR scale showed both feasibility and internal consistency. The results obtained using Snail-KR were found in Laupichler et al. [6]excludes seven items from the original version. Nevertheless, further verification with a larger, more diverse sample is required to confirm its reliability and applicability.
