Treatment planning is a crucial step in the workflow of radiotherapy. With the advancement of computer technology, complex and challenging radiotherapy plans can now be created while better preserving normal tissues. However, in such cases, planners previously required hours, even days, of trial and error to optimize difficult treatment plans. Recently, artificial intelligence has been utilized to automate and enhance various aspects of medical science. Artificial intelligence can personalize the entire therapy process, leading to improved treatment outcomes and ultimately assisting people, especially when considering time savings. Furthermore, artificial intelligence can expedite repetitive tasks, benefiting daily clinical routines. This can alleviate the burden on clinical staff, increase efficiency, and help reduce costs. Numerous algorithms have been developed to better support planners in radiotherapy treatment planning. These algorithms have already had a significant impact on automating the planning process and improving the consistency and quality of treatment plans. The purpose of this study is to review new artificial intelligence-based treatment planning applications, such as deep learning algorithms and emerging research directions, and provide information on the challenges of artificial intelligence-based treatment planning and future work.