My thesis explores the use of reinforcement learning to develop intelligent agents in a basic 3D tank battle environment. The project's main goal is to train an AI agent to move around the game space, aim, and attack using the Proximal Policy Optimization algorithm. I used Unity to create the project and set up the ML-Agents Toolkit for training. The AI agent demonstrated basic tactical behaviors such as seeking enemies and avoiding obstacles. Overall, the thesis highlights the importance of reinforcement learning in game development.