This paper introduces the common AI methods Finite State Machine and Fuzzy Logic. First the theoretical basics of the methods are explained and later an example shows how it is possible to implement these theories in a mobile game. The implemented concept of the AI opponent also covers how to model different skill levels for the AI. In addition there are adaptive methods described how the skill level can be changed during a game. Based on these methods an alternative method is created in which the AI skill searches and adapts to the skill of the human player. This adaptive algorithm should achieve that the challenge level of the AI matches the skill level of the human player. In the end the implemented method is tested. The different AI skill levels play numerous times against each other. Half of the matches are played without the adaptive skill method and the other half is played with the method. This test shows the efficiency of the created method.