This study examines the adoption, strategies, and challenges of Artificial Intelligence (AI) in higher education across the Middle East and North Africa (MENA) region, based on data from 508 participants at 259 institutions across 19 countries. Using a mixed-methods approach, it captures quantitative and qualitative insights into AI integration, barriers, and opportunities for development. Guided by the Technology Acceptance Model (TAM) and Diffusion of Innovations (DOI) theory, the findings reveal key disparities in AI adoption across economic groups, with high-income and lower-middle-income countries demonstrating significantly greater integration in administrative and educational functions. AI adoption is still in early stages for 43.9% of institutions, with Adaptive Learning Platforms (ALP) (29.6%) and AI-enhanced research tools (26.9%) being the most commonly implemented technologies. Barriers, particularly in low-income countries, include financial constraints (17.9%), infrastructure issues (17.1%), and unclear policies. Significant correlations highlight the positive relationship between AI integration in administrative functions and improvements in student support and institutional planning (Spearman's rho = 0.773 and 0.754, p < 0.001), as well as between AI-driven personalization and student engagement (Spearman's rho = 0.876, p < 0.001). The study underscores the need for targeted investments, faculty training, policy development, and regional collaboration to accelerate AI adoption and maximize its benefits across MENA higher education institutions.