Nowadays, the size of data recorded and stored in enterprises information systems (IS) is increasing every second. To face to this phenomenon, contemporary technologies play a major role for gathering, analyzing, storing, and distributing data that enables organizations to make smart decisions and to take full control of their activities. The traditional Business Process (BP) mining techniques were intensively used to discover, monitor, and optimize processes from event-logs without needing any priory model. However, the majority of the suggested algorithms have exhibited their limits (such as discovering nested loops, managing duplicate and hidden tasks as well as dealing with concurrent processes). In parallel, recent advances in the Artificial Intelligence (AI) discipline have generated a great deal of enthusiasm in a large spectrum of research area. In this perspective, AI methods emerge as one of the pillars to overcome the drawbacks of the conventional approaches allowing anomalies detection, prediction and recommendation tasks on ongoing process instances at run-time. The aim of this work is to explore towards the use of AI techniques in the field of business process mining by presenting a state-of-the-art review ranging from traditional PM approaches to AI ones, as well as outlining a prospective road-map for mining business process models basing on AI techniques.