Knowledge, Attitudes, and Practices of Artificial Intelligence in Dentistry: A cross-sectional survey

Introduction

The Fourth Industrial Revolution has reshaped how humans live, work, and interact, with technological advances rapidly influencing multiple sectors.¹ Among these, Artificial Intelligence (AI) and Augmented Intelligence (AUI) have drawn significant attention and are now increasingly integrated into healthcare and higher education.² AI can be broadly described as a computer-based simulation of human cognitive abilities, capable of performing tasks such as speech recognition, natural language comprehension, and complex decision-making.³ The term artificial intelligence was first introduced by John McCarthy at Dartmouth University in 1956, and since then, AI has been widely regarded as a transformative force across diverse domains, including supply chain management, transportation, and healthcare.⁴

The rapid rise of AI applications in the last decade can be attributed to breakthroughs in advanced algorithms, cost-efficient computing resources such as Graphic Processing Units (GPUs), and the availability of extensive annotated datasets.⁵ Dentistry, in particular, has recently witnessed significant advances through AI-driven technologies. Following the basic computational framework of input, processing, and output, AI systems can process diverse data types in dentistry ranging from auditory data (e.g., handpiece sounds) and textual information (e.g., patient records, treatment parameters) to image-based data (e.g., radiographs and clinical photographs).⁶ These inputs, when processed through neural networks, yield outputs such as diagnostic insights, prognoses, treatment planning, or disease prediction.⁷ AI models, including conventional neural networks and advanced deep learning approaches, have already been applied in root canal anatomy analysis, staging of malignant lesions, detection of proximal caries, Computer Aided Design and Computer Aided Manufacturing (CAD/CAM)-based prosthesis design, and dental implant placement.⁸

AI has become an essential component of dental healthcare education. In both preclinical and clinical settings, AI tools enhance learning experiences by offering adaptive, personalized, and mobile-based education.⁹ AI systems can tailor learning resources to individual student needs, address knowledge gaps, and provide real-time feedback.¹⁰ Furthermore, AI facilitates the continuous monitoring of academic progress, clinical exposure, and professional development while supporting mentorship and career guidance. These features have positioned AI as an invaluable resource in competency-based education, which is becoming increasingly relevant in global dental curricula.¹¹

Despite its potential, AI integration remains uneven across regions. While Western countries such as United States have incorporated AI into healthcare education and practice for decision support,¹² developing nations face infrastructural, financial, and sociocultural barriers. For example, India’s healthcare sector struggles with challenges such as limited institutional investment, resistance from healthcare providers, inadequate numbers of AI-trained professionals, and ethical concerns surrounding patient data security.^13,14 Additionally, issues such as medicolegal implications, public perception, and fears of physician replacement further hinder the acceptance of AI in medical and dental practice. Addressing these challenges will require targeted investments, robust data policies, faculty training programs, and collaborative research across global institutions.¹⁵

Although AI applications in dentistry are expanding, there remains a paucity of research exploring the knowledge, attitudes, and practices (KAP) of dental students regarding AI in India. Understanding how undergraduate and postgraduate dental students perceive and engage with AI is crucial for developing effective curricula, improving clinical training, and fostering responsible adoption of these technologies. This research aims to assess the level of awareness, attitudes, and readiness for AI integration among dental students, thereby contributing evidence to guide educational policies and infrastructure development in India and other developing countries.

Methods

Results

Out of 254 dental students invited to participate, 208 completed the survey, resulting in a response rate of 81.8%. The average age of participants was 24.29 ± 2.48 years, ranging from 21 to 45 years. The group consisted of 146 females (70.2%) and 62 males (29.8%). Among them, 120 were undergraduate (UG) students and 88 were postgraduate (PG) students. Specifically, the sample included 44 (36.5%) final-year BDS students, 76 (21.2%) house surgeons, 34 (16.3%) first-year MDS students, 30 (14.4%) second-year MDS students, and 24 (11.5%) third-year MDS students [ Table 1].

Discussion

The key findings of this study indicate that most students were familiar with the term AI, postgraduate students had greater awareness of AI applications in medical and dental fields, a considerable proportion of undergraduate students found AI easier to apply compared to postgraduates, and the majority of respondents emphasized the indispensable role of clinicians in AI-assisted dentistry. These outcomes suggest that dental students remain in an exploratory phase, with a degree of uncertainty about adopting AI applications in their field.

The gender distribution of participants in this survey (70.2% female, 29.8% male) aligns with previous reports by Murali et al.¹⁷ (73% female, 27% male) and Elhijazi et al.¹⁸ (71.4% female, 28.9% male). The highest participation was from interns (36.5%), while final-year MDS students were least represented (11.5%). By contrast, Murali et al.¹⁷ reported lower participation from interns (19.3%) and higher involvement of postgraduates (20.7%), whereas Elchaghaby et al. found most respondents were fifth-year students (53%) and fewest were in their third year (21%).¹⁹

In the present study, 95.6% of participants had basic knowledge of AI, closely matching Murali et al.’s findings (94%). Awareness of machine learning and deep learning was higher among PG students (82.9%), likely due to the integration of AI into clinical practice and increased exploration of new technologies through social media and smartphones. Supporting this, Yüzbaşıoğlu²⁰ reported that 76.1% of students learned about AI from social media, while Aldowah et al.²¹ found similar results (78%). Compared with medical students in Syria¹⁶ (34.7%) and Pakistan²² (35.3%), our respondents demonstrated higher knowledge of AI subtypes. However, only 25.8% of UG students knew about AI’s applications in dentistry, versus 90.9% of PG students who were aware of its role in healthcare. This disparity reflects the limited academic and clinical exposure of undergraduates to AI, underscoring the need for curriculum integration. Medical studies from Syria¹⁶ (87.4%), Pakistan²⁰ (74.4%), and the UK²³ (78%) further reinforce AI’s perceived importance in healthcare. Similarly, 77.3% of our PG students acknowledged its role in oral radiology, consistent with reports from Pakistan²² (74%) and Syria¹⁶ (73%).

Regarding future implications, 33.4% of UG and 43.2% of PG students believed AI might replace dentists. These findings contrast with Jeong’s study²⁴ where 72% disagreed with this notion, but align with Elhijazi¹⁸ where 53.7% considered replacement possible. The higher proportion in our study may stem from misconceptions about AI’s scope and limitations in dentistry. Nonetheless, most participants (77.8%) agreed that AI supports early diagnosis and disease severity assessment, echoing global trends. AI systems have already demonstrated their value in enhancing diagnostic accuracy and reducing human error. Additionally, more than half of respondents (57.7%) acknowledged AI’s importance in oral surgery, a view shared by students in Turkey,²⁰ Saudi Arabia,²⁵ and India.²⁶

Interestingly, 36% of students felt AI could become a burden for practitioners. While AI is revolutionizing decision-making and personalized treatment, its integration also raises concerns regarding job displacement, information overload, and potential erosion of clinical skills if overreliance occurs. To prevent disruption, AI should be positioned as a supportive tool rather than a substitute in oral healthcare.

Overall, the findings show moderate baseline knowledge of AI, broad acceptance of its inclusion in the curriculum, and a generally positive attitude toward its use in dentistry. Students also demonstrated enthusiasm for future learning. Therefore, we recommend hands on workshops, seminars, and training programs on machine learning and deep learning in clinical dentistry. This would enable future dentists to harness AI effectively, reducing errors and enhancing treatment quality. Addressing identified knowledge gaps will require expanding AI training within dental education to eliminate misconceptions and prepare these students for the future of digital dentistry.

Conclusion

A significant proportion of dental students in this study demonstrated basic knowledge of AI and a positive attitude toward incorporating it into the dental curriculum. Students recognized its value in dental applications and expressed interest in future learning. Universities and policymakers should collaborate to integrate AI training into dental education, equipping future dentists to utilize AI in routine practice and ultimately improve the quality of dental healthcare.

Ethics and consent

The protocol was approved by the Institutional Review Board of Bapuji Dental College and Hospital, India (Approval number: ECR/1652/Inst/KA/2022/24-04/08-007). The Committee on Research Ethics followed internationally recognized guidelines for the protection of human subjects, including the Declaration of Helsinki, the Belmont Report, and CIOMS principles. Prior to participation, all individuals were provided with a clear explanation of the study and gave informed consent electronically. The consent form described the study’s purpose, estimated completion time, and was written in simple, accessible language. Participation was entirely voluntary, and responses were collected anonymously. Participants were informed of their right to withdraw at any point, and assured that their personal information would remain confidential. Access to the dataset was restricted to the principal investigator.