The review searched PubMed and Scopus for studies published between November 2014 and March 2026 and included 42 publications. Applications covered teaching and learning, assessment and academic-performance prediction, including chatbots, simulation, generative models, radiograph interpretation, writing support and early-risk prediction. Several studies reported improved engagement or efficiency and moderate-to-high agreement with human evaluators.
Key findings
- AI may support personalised and self-directed learning, clinical reasoning, assessment consistency and earlier academic support. Agreement with human graders does not establish that AI can replace educators, and performance in individual tasks does not guarantee transfer across curricula or institutions.
Why this matters globally
Dental schools worldwide face common questions about accuracy, fairness, privacy and the competencies students need. The review helps define a curriculum agenda, while stronger global evidence will require multicentre evaluation, error reporting and measurement of real clinical competence.
Thai researcher contribution
Chulalongkorn University-affiliated authors contributed the evidence synthesis and its interpretation for dental education, positioning a Thai academic team within the international debate on responsible AI adoption.
Limitations to consider
The included studies and AI tools are heterogeneous and rapidly evolving. Many may be small or context-specific. Without pooled effects and a clear quality appraisal, the review cannot show that every AI application consistently improves learning or assessment accuracy.