Motion planning and predicting the future states of the surrounding environment are among the main challenges in automated driving. In lane change and merge maneuvers, it is important to know how neighboring vehicles will react in the imminent future. Such a problem becomes more demanding in the absence of inter-vehicular communication (such as V2V, V2X, etc.). Human driver models, probabilistic approaches, rule-based techniques, and machine learning methods have addressed this problem only partially as they do not focus on the behavioral features of the vehicles. In addition, the framework that undertakes the prediction is expected to be fast in providing the path planner with the estimate of future states of the vehicles. Constructing such a fast structure, con-sidering interactions between vehicles, is the main motivation of this study. In this paper we present a fast receding horizon algorithm based on Monte Carlo tree search for real-time path planning in highway scenarios. Inspired by recent results in [1] [2], we adopt a level-k game framework for predicting the strategy of the neighboring vehicles. Our simulations show promising results with fast computations.