Emergency call management faces increasing pressure due to rising call volumes, aging populations, and the necessity for cost-efficient responses. This paper introduces the IAppel project, which focuses on integrating Generative Artificial Intelligence into emergency call processing through decision trees. Unlike traditional call-handling methods, this AI-assisted system utilizes advanced large language models to classify and route emergency calls efficiently. The system processes audio data using automatic speech recognition for transcription, employs structured prompts and decision trees to accurately categorize calls by urgency and incident type, and subsequently directs them to the appropriate emergency services. An evaluation based on real-world emergency call transcripts demonstrated high accuracy, with precision, recall, and F1-score metrics indicating strong alignment between AI predictions and human operator decisions. While the AI models successfully mirrored human performance, they also uncovered cases where human operators had misclassified incidents, highlighting AI’s utility as a decision-support tool.