This study investigates the introduction of an Artificial Intelligence (AI)-based collision avoidance system to prevent collisions between merchant ships and fishing vessels. Through the analysis of overseas case studies and a review of the necessary technical elements, the study proposes specific implementation measures. An analysis of maritime accident statistics for merchant ships over the past decade revealed that approximately 87% of accidents were caused by human error. Existing collision prevention measures have mainly focused on crew training and adherence to navigation rules, but they have limitations in completely eliminating human error. Therefore, the introduction of an AI-based collision avoidance system is seen as an effective solution for enhancing maritime safety. This study examines the technical concepts and key elements of AI-based collision avoidance systems and verifies the system's effectiveness through international case studies. Additionally, the study analyzes the current state of the Korean shipping industry and suggests procedures and methods for implementation, evaluating the positive impacts of technology adoption. The results confirm that AI-based collision avoidance systems can significantly reduce collision risks through real-time data analysis and autonomous avoidance functions, as demonstrated by overseas case studies. However, there are various challenges to implementation, including the development of technology, legal regulation updates, and strengthening industry-academia collaboration. Policy support is emphasized as necessary to address these challenges. This study aims to contribute to the academic and practical development of the field by proposing practical measures that can contribute to maritime safety, including the prevention of ship collisions.