The autonomously operating vehicle (AV) is revolutionizing the transportation space rapidly, driven by AI inside it. This paper looks at the recent work done on the AI algorithms making self-driving cars work: perception, path planning, and decision-making. Using AI, vehicles can interpret their environment by reading data from sensors like cameras, LiDAR, and RADAR using deep learning and computer vision techniques. With all these, autonomous vehicles are now able to detect objects and predict movements in addition to making decisions critical for safe navigation in real time. On this part, the artificial intelligence algorithms have advanced on path planning to handle static and dynamic environments so that self-driving cars can take the most complex routes without the encumbrance of bumping into obstacles (Lee and Patel in Journal of Machine Learning and Applications 9:450–465, 2022). Graph search algorithms are essential in this aspect, including Dijkstra’s and A* in finding the safest and most efficient routes. Decision-making processes the handling of ethical dilemmas and real-time risk assessment via AI is also under observation. This again renders importance to the machine learning models in uncertain traffic conditions. Despite all the progress that has been made, however, AI-powered autonomous driving systems face tremendous difficulty. Among those issues are safety, reliability, and great computational power required for near-time processing. Regulatory hurdles and ethical considerations in AI decision-making become more challenges on the road. Conclusion on a closing note, this paper discusses recent innovations in the form of improvement in neural networks and edge computing, exploring how fully autonomous driving will look in the future.