<p>The global maritime sector is undergoing a profound transformation as seaports evolve from traditional cargo hubs into intelligent, sustainable ecosystems. This paper examines the strategic integration of Artificial Intelligence (AI)-enabled Digital Twins (DTs) as a foundational framework for achieving carbon-neutral smart seaport operations. Drawing on recent theoretical developments and empirical case studies, the study traces the conceptual evolution of digital twin architecture in maritime logistics from perception and transmission layers to AI-driven intelligence and application interfaces. Advanced AI methodologies, including Long Short-Term Memory (LSTM), Random Forest (RF), and Reinforcement Learning (RL), are analyzed for their roles in predictive scheduling, risk classification, and dynamic resource optimization. The paper also explores how AI-DT integration enables proactive decarbonization strategies, including Just-in-Time vessel arrivals, real-time emissions mapping, and adaptive energy management. Situated within the Industry 5.0 paradigm, the study highlights a critical shift from efficiency-centric automation toward system-wide antifragility, human-centricity, and environmental stewardship. Findings indicate measurable improvements across key sustainability metrics, including reductions in carbon emissions and vessel turnaround times. This work contributes to the growing body of knowledge on smart port governance and offers strategic insights for port authorities, logistics planners, and policymakers navigating the transition to green maritime infrastructure.</p>