The next telecommunications standard, 5G, envisions that the future networks will support advanced use cases, such as Internet of things while supporting voluminous simultaneous connections with high bandwidth as well as low latency. Further, these 5G deployments will not be static in nature, with new use cases and service requirements evolving in future. Such requirements pose many deployment and operational challenges to MNOs. These use cases would not only require the networks to be aware of connectivity related parameters, but also adapt intelligently based on parameters beyond the network. This requires the 5G networks to be capable of addressing conditions which are not foreseen at the time of designing them. Such capability requirements can be adequately addressed by advances in the field of AI and machine learning. The objective of this paper is to explore ways to leverage AI and machine learning for enhancing the 5G network deployments and operations. This paper attempts to decipher future demands from the 5G networks analyzing specific requirements in the areas of network planning, network operations and network optimization. This paper also discusses the strategic perspective for MNOs to benefit from applications of AI in 5G networks.