Barriers to and Needs for the Implementation of Artificial Intelligence in Radiation Therapy Planning in Saudi Arabia. A Cross-Sectional Survey

Objectives: To assess radiation oncology physicians' perceptions of artificial intelligence (AI) in radiation therapy planning in Saudi Arabia and identify barriers to and support needs for its implementation.Methods: A cross-sectional online survey was distributed through the Saudi Association of Radiation Oncology email list to 142 physicians from April to May 2026. We analyzed responses using descriptive statistics and exploratory 2-sided Fisher exact tests comparing residents with nonresidents.Results: Forty-three complete responses were analyzed (response rate: 30.3%). A total of 33 (76.7%) respondents agreed or strongly agreed that AI would improve planning accuracy and efficiency. Nineteen (44.2%) reported no workplace access to AI tools, and 34 (79.1%) reported no formal AI training. The most common concerns regarding AI were output accuracy or validation (n = 28, 65.1%), loss of clinical judgment (n = 22, 51.2%), and legal liability (n = 21, 48.8%). The most requested forms of support were hands-on training or simulation (n = 35, 81.4%), regulatory and legal guidance (n = 30, 69.8%), and vendor technical support (n = 29, 67.4%). Nonresidents more often reported practical implementation barriers, including funding limitations, accuracy issues, and inadequate support; however, these comparisons were confounded by age and experience differences and should remain exploratory and hypothesis-generating.Conclusion: Respondents viewed AI favorably, but gaps regarding access, training, validation, funding, and governance may limit safe implementation. Structured education, local validation, quality assurance, technical support, and clear institutional and regulatory guidance are needed.