Assessment of Perceptions, Trust, and Adoption of Clinical Artificial Intelligence Tools Among Physicians and Dentists, Riyadh, Saudi Arabia.

PURPOSE: Clinical artificial intelligence (AI) tools are increasingly integrated into healthcare; however, their acceptance varies across professional groups. Understanding clinicians' familiarity with and perceptions, trust, and adoption of clinical AI tools is essential for safe and effective implementation. This study assessed physicians' and dentists' perceptions, trust, and intention to adopt clinical AI tools in governmental university hospitals in Riyadh, Saudi Arabia.

PARTICIPANTS AND

METHODS

A cross-sectional survey was conducted among 400 clinicians selected from three major governmental university hospitals in Riyadh (King Saud University Medical City, King Abdulaziz Medical City-National Guard, and King Abdullah Bin Abdulaziz University Hospital), which serve as large tertiary academic centers at the forefront of digital health and AI implementation in Saudi Arabia. Data were collected using a structured online questionnaire assessing familiarity, usage patterns, perceptions, trust, and adoption intention. The instrument demonstrated high reliability across subscales (Cronbach's α: 0.880-0.952). Data were analyzed using descriptive statistics, non-parametric tests, and multivariate linear regression.

RESULTS

Familiarity with AI was high (89.5%), with physicians reporting greater familiarity than dentists. Overall perceptions were moderate (mean 3.2-3.8 on a 5-point scale), with physicians showing more positive views. Trust was generally low (mean 3.1-3.6), particularly among physicians; dentists reported moderate trust. Intention to adopt was moderate (mean 3.5-3.9), with physicians demonstrating higher adoption intention than dentists. Residents, physicians, and those in digitally advanced hospitals had significantly higher perceptions, trust, and adoption intention scores (p <0.05). Multivariate regression identified work setting, resident status, and physician specialty as the strongest predictors of higher scores across all domains.

CONCLUSION

Despite high awareness, perceptions of clinical AI tools and intention to adopt them remain moderate and constrained by low trust. Targeted training, transparent systems, and regulatory support are essential for enhancing integration into practice.