Are ChatGPT, My AI Snapchat, and Metaverse used by dental students as reliable sources of dental education?

A profound understanding of dental students' use of artificial intelligence (AI) applications and their perceptions is essential for promoting responsible adoption and guiding integration into dental education.

This study evaluated dental students' perceptions, usage patterns, and trust in ChatGPT, Snapchat's My AI, and the Metaverse as educational tools.

A cross-sectional online survey was administered to undergraduate dental students at RAK College of Dental Sciences, United Arab Emirates, between February and May 2024. The questionnaire comprised 29 questions organized into five sections. The questions were formulated as multiple-choice questions. Descriptive statistics were calculated. Pearson's chi-square tests with post-hoc adjusted standardized residuals (ASRs) and Bonferroni corrections examined categorical associations. Ordinal logistic regression assessed predictors of AI awareness, and paired-samples t-tests with Cohen's d compared trust in ChatGPT vs. Snapchat AI.

The response rate was 57%, of which 70% were females. Self-rated awareness of AI increased across academic years, with fifth-year students rating their awareness score as 4 ± 0.9 out of 5, compared to 3 ± 1.4 in the first year. Ordinal regression showed no significant effects of year, gender, or their interaction (p > 0.19). ChatGPT was the most used tool (81.5%), followed by grammar correction tools (75%) and Snapchat AI (74.4%), while Metaverse use was limited (28.6%). Chi-square analyses confirmed significantly greater use of ChatGPT, Grammarly, and Snapchat AI compared with Metaverse (p < 0.001). Fourth-year students most often used AI for academic or clinical purposes (41.7%). Educational potential was endorsed by 78% of students, while privacy and data security were the predominant concerns (78%). Compared with Snapchat AI, ChatGPT was significantly more often used for education, preferred for quick responses, and more frequently associated with positive beliefs about future learning. Paired-samples t-tests demonstrated consistently higher trust in ChatGPT across all academic years, with moderate-to-large effect sizes (Cohen's d = 0.67-0.91; p < 0.01).

Dental students reported widespread adoption and higher trust in ChatGPT compared to other AI tools. While recognizing its educational potential, concerns about accuracy and privacy underscore the need for integrating AI literacy and evidence-based evaluation into dental curricula.