Adrenal tumors are highly heterogeneous endocrine disorders with markedly varied prognoses across clinical subtypes, underscoring the critical need for precise diagnosis and treatment. Artificial intelligence (AI) is increasingly employed in this field to improve diagnostic accuracy, refine prognostic assessment, and advance personalized medicine. This review systematically summarizes recent advancements in AI for the precise diagnosis, treatment planning, and prognosis prediction of adrenal tumors, while examining its strengths and limitations. AI has significantly enhanced precision in managing adrenal tumors. It achieves an area under the curve (AUC) of 94.5% for subtype classification, 93.7% for predicting intraoperative hemodynamic risk, and 91.7% sensitivity for tumor metastasis, and 95.5% specificity for prognosis assessment. Additionally, AI effectively supports surgical decision-making and personalized medication recommendations. However, current applications face several challenges, including high heterogeneity in single-center data, insufficient algorithm interpretability, and inadequate integration of multimodal and multi-omics data. Future efforts should focus on establishing standardized multicenter databases, developing explainable AI techniques, enhancing multidimensional data integration, and conducting large-scale clinical validation. These measures are essential to accelerate the clinical translation of AI and to propel the field toward more precise and individualized patient management.