Vehicular Ad-Hoc Networks (VANET) have gained significant attention in recent years as a promising technology for enabling intelligent transportation systems and smart cities. VANET enables vehicles to communicate with each other and with roadside infrastructure to exchange information about traffic, road conditions, and other relevant data. However, VANET faces several challenges such as high mobility, varying network conditions, and congestion, which traditional routing and mobility protocols may not be able to handle efficiently. Therefore, there is a need for novel approaches to improve the performance of VANET.In recent years; AI-based approaches for Routing Protocols and Mobility Framework have been proposed to address the limitations of traditional protocols. AI-based approaches like Reinforcement Learning, Neural Network, Swarm Intelligence, Deep Reinforcement Learning, Fuzzy Logic, and Genetic Algorithm have shown promising results in improving the performance of VANET. However, the implementation of these approaches in VANET requires significant computational resources, training data, and may be more complex than traditional protocols, which presents practical challenges.In this review, we provide a comprehensive overview of the different routing protocols and mobility frameworks for VANET, as well as the AI-based approaches that have been proposed to improve their performance. We analyze the strengths and weaknesses of each approach and compare them with traditional protocols like AODV. Our analysis highlights the potential of AI-based approaches to overcome the limitations of traditional protocols and improve the performance of VANET. However, we also identify practical challenges that need to be addressed for the implementation of AI-based approaches in VANET.