Voltage stability is referred to a fundamental requirement for stable and reliable operation of modern power systems, particularly under growing penetration of renewable energy sources, such like; distributed generation, and inverter-based technologies. This paper reviews recent advances of three years in voltage stability assessment and improvement, give emphasis to the integration of artificial intelligence (AI) techniques and metaheuristic optimization algorithms. It look at data-driven approaches for predictive modeling and real-time decision-making, together with optimization-based strategies for reactive power dispatch, flexible AC transmission system (FACTS) device placement, and distributed generation siting. In generalit used voltage stability indices, optimization objectives, and constraints are formalized, with add-ontables. Comparative analyses summarize the performance of AI models, metaheuristic methods, and hybrid frameworks. Key challenges are discussed, including generalization under topology changes, Measurementsparsity, renewable energy uncertainty, and the need for explainable decision support. Future directions highlight physics-informed AI, graph-based learning, robust optimization, and coordinated multi-agent control for sustainable and resilient voltage stability management