Unmanned Traffic Management (UTM) systems rely on humans in the loop for reliable and accountable decision-making. Typically, system operators face significant cognitive load due to the high volume of information, the need for multitasking, and real-time problem-solving to ensure safe and efficient airspace management. These demands can lead to cognitive fatigue and an increased risk of errors. Due to their extensive attention abilities, we propose integrating large language models (LLMs) into UTM systems as decision-assistance tools. LLMs can perform real-time data analysis, interpret situations, and suggest optimal solutions, which helps UTM system operators make more reliable, informed, and timely decisions. This poster shows a use case where we integrate an LLM into an airspace monitoring and surveillance system. This is to verify UAVs’ compliance with remote identification regulations and conformance to mission plans to warn the system operator of any violations. We validate the solution by evaluating the accuracy of the LLM responses and the time needed to generate them for different zero-shot prompts in three scenarios. The results show that the LLM can generate helpful messages to the system operator with an accuracy of up to 91.7% within 5.7 s on average. In future work, LLMs will be integrated into other UTM systems. The communication between these LLMs will be enabled and evaluated towards a secured, fully GenAI-empowered UTM with zero trust.