The descriptive numbers point toward growing familiarity with AI tools. The share of students reporting frequent AI use rose from 22 of 109 students, or 20.2 percent, in 2025 to 27 of 92 students, or 29.3 percent, in 2026, while the proportion reporting they rarely or never used AI fell from 35 of 109 students, or 32.1 percent, to 19 of 92 students, or 20.7 percent. Descriptively, the 2026 cohort also more often identified treatment planning as the principal perceived benefit of AI. These are self-reported perceptions from two separate samples, however, not objective measurements of tool adoption, and the authors treat them as cohort-level observations rather than confirmed trends.

The statistical story is more cautious than the raw percentages suggest. When the researchers compared AI-use frequency between cohorts using a Mann–Whitney U test, the unadjusted difference reached significance (U = 4243.5; unadjusted P = .042; r = .14), and the distribution of perceived benefits also differed in unadjusted analysis (χ²₃ = 8.27; unadjusted P = .041; Cramér’s V = 0.20). Yet both effect sizes were small, and because the study prespecified no primary outcome, all inferential analyses were treated as exploratory. After applying the Holm step-down procedure to correct for multiple testing, neither comparison remained statistically significant, with adjusted P values of approximately .82 for both. The authors are explicit that these small cohort-level differences should be interpreted as exploratory rather than confirmatory evidence of temporal change.

What the data do show consistently across both cohorts are the training gaps. Familiarity with regulations and legal guidelines governing AI remained limited: 68 of 109 students, or 62.4 percent, in the 2025 cohort and 46 of 92 students, or 50 percent, in the 2026 cohort reported being not familiar with such guidelines. Perceived institutional preparedness was similarly low and essentially unchanged between years, with 69 of 109 students, or 63.3 percent, in 2025 and 62 of 92 students, or 67 percent, in 2026 reporting that their institution did not provide sufficient AI-related training. The authors conclude that these ethical, legal, and institutional training gaps evident in both cohorts support the need for structured AI literacy, regulatory awareness, and clinically oriented AI training within dental curricula in Bulgaria.

The study’s design shapes how its findings should be read. Two anonymous cross-sectional questionnaire surveys were administered to independent cohorts, meaning the same students were not followed over time; each year’s third-year class was sampled separately. Demographic variables collected included gender, age, and country of origin, a detail that matters because the English-language program enrolls international students, a group the authors note remains insufficiently explored in AI-related research. Categorical variables were compared with chi-square tests and ordinal responses with Mann–Whitney U tests, with statistical interpretation based on Holm-adjusted P values. Because usage frequency was measured by student report rather than objective logs, the rise from roughly one in five to nearly one in three students reporting frequent AI use reflects perceived behavior, not verified adoption. The authors note that AI is increasingly influencing dental education, clinical decision-making, diagnostics, and treatment planning, and that cohort-level differences in AI-related perceptions and use among dental students from consecutive academic years remain insufficiently explored. Within that scope, the survey provides a like-for-like snapshot of two consecutive cohorts, and its central, statistically robust observation is not a confirmed rise in AI use but the persistence of regulatory and training gaps across both years.

Subject of Research: Dental students' perceptions and use of artificial intelligence in dental education in Bulgaria

Article Title: Dental students’ perceptions and use of artificial intelligence in dental education in Bulgaria: a comparative cross-sectional survey of two independent cohorts in 2025 and 2026

Article References: Stefanova, V., & Zhekov, K. (2026). Dental students’ perceptions and use of artificial intelligence in dental education in Bulgaria: a comparative cross-sectional survey of two independent cohorts in 2025 and 2026. BMC Medical Education. https://doi.org/10.1186/s12909-026-10304-9

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10304-9

Keywords: artificial intelligence, dental education, dental students, Bulgaria, cross-sectional survey, AI literacy, regulatory awareness, treatment planning, medical education, cohort comparison, Dental, students