The exponential increase in vehicles, rapid urbanization, and rising transportation demands have created unprecedented challenges for urban traffic management worldwide. Traditional fixed-time traffic signal systems have proven inadequate, with studies indicating they account for approximately 10% of global traffic delays and contribute to millions of hours lost in congestion annually. This comprehensive literature review examines the emerging field of intelligent traffic management systems that leverage artificial intelligence techniques to optimize traffic light control. The review categorizes and analyzes various AI approaches including Fuzzy Logic (FL), Metaheuristic algorithms (MH), Dynamic Programming (DP), Reinforcement Learning (RL), Deep Reinforcement Learning (DRL), and hybrid techniques. We explore both single-intersection and multiple-intersection implementations, examining their architectures, methodologies, and performance metrics. Simulation results from multiple studies demonstrate significant improvements in key performance indicators, with some implementations reducing queue lengths by 28%, average vehicle latency by 27%, and CO₂ emissions by 28%. The review also discusses microsimulation tools essential for evaluating these systems, including CORSIM, AIMSUN, VISSIM, and SUMO. Despite promising results, challenges remain in scalability, real-time implementation, and integration with existing infrastructure. Future research directions point toward integration with connected and autonomous vehicles, multi-modal traffic management, and smart city ecosystems.