Navigating the open seas is a complex task, demanding adherence to the International Regulations for Preventing Collisions at Sea (COLREGs). These rules, designed for human interpretation, present a unique challenge for autonomous surface vehicles (ASVs). How can we ensure these vessels not only avoid collisions but also understand and follow the nuanced rules of maritime navigation? Researchers are exploring the potential of Large Language Models (LLMs), like those powering ChatGPT, to tackle this challenge. LLMs, known for their ability to process and understand natural language, offer a promising approach to deciphering the complexities of COLREGs. This research introduces a novel system that integrates an LLM with an ASV’s control system. The LLM acts as a high-level decision-maker, interpreting real-time data like vessel headings, proximity, and potential collision risks. It then translates this information into actionable commands, such as 'give-way' or 'stand-on,' while providing clear explanations for its decisions. Imagine an AI captain that not only steers the ship but also explains its actions based on maritime law! Tested in simulated scenarios involving overtaking, head-on, and crossing situations, the system demonstrated a remarkable ability to adhere to COLREGs. This included making correct judgments based on relative vessel positions and potential risks, all while providing transparent reasoning. While the research shows great promise, challenges remain. Ensuring the system's reliability in real-world, unpredictable conditions is crucial. Future research will likely focus on refining the system, rigorously testing it, and ultimately building trust in AI's ability to navigate safely and responsibly. This innovative approach represents a significant step towards realizing truly autonomous maritime navigation, paving the way for safer and more efficient seafaring in the future.