Acceptance and Readiness for AI Among United Arab Emirates–Based Health Care Practitioners: Exploratory Cross-Sectional Survey

Background: Artificial intelligence (AI) can enhance diagnostic accuracy, efficiency, and decision-making in health care, but real-world impact depends on practitioners' acceptance and readiness to use AI in clinical workflows. The United Arab Emirates offers a policy-driven context to study these factors, given active national AI strategies and rapid health system digitization. Objective: This study aimed to develop and validate a model explaining how trust, perceptions, perceived risk, and perceived benefit shape practitioners' acceptance of AI and, in turn, their readiness to implement AI in clinical practice. The model integrates the Technology Acceptance Model, the Unified Theory of Acceptance and Use of Technology, and the Theory of Trust and Acceptance of Artificial Intelligence Technology. Methods: We conducted a cross-sectional online survey of 182 United Arab Emirates-based health care practitioners (physicians, nurses, dentists, and allied health staff). Constructs included trust, perception, perceived risk, perceived benefit, acceptance, and readiness. Knowledge of AI was also assessed using true or false statements. We performed confirmatory factor analysis and structural equation modeling, reporting standard fit indices. The survey adhered to the Checklist for Reporting Results of Internet E-Surveys guidelines, and ethics approval and electronic consent were obtained. Results: Trust was positively associated with perception (β=.704; P<.001) and perceived benefit (β=.191; P=.02) and negatively associated with perceived risk (β=-.301; P<.001). Acceptance was positively associated with trust (β=.452; P<.001), perception (β=.459; P<.001), and perceived benefit (β=.168; P=.002), and negatively associated with perceived risk (β=-.140; P=.009). Acceptance strongly predicted readiness (β=.874; P<.001). The model fit indices are standardized root-mean-square residual of 0.068, root-mean-square error of approximation of 0.0913, goodness-of-fit index of 0.802, adjusted goodness-of-fit index of 0.763, and comparative fit index of 0.906. Our knowledge assessment found notable gaps among participants, underscoring a need for education and training. Our study sample was predominantly drawn from Dubai-based health care settings (103/182, 57%) and nursing roles (71/182, 39%); therefore, these findings primarily reflect the Dubai health regulatory environment and nursing workflows and may not generalize to the broader federal health care system across all Emirates. Conclusions: Trust is a central lever for advancing AI acceptance and implementation readiness among the study cohort of United Arab Emirates-based health care practitioners. Implementation programs should prioritize building institutional and technical trust (transparency, safety, and governance), reducing perceived risk (privacy, security, and reliability), and amplifying perceived benefits through hands-on demonstrations and workflow-aligned use cases. Targeted training to close knowledge gaps should accompany policy and organizational measures aligned with national AI strategies to accelerate responsible, clinician-in-the-loop adoption.