Researchers at George Washington University have developed a system to automatically summarize financial news using Large Language Models (LLMs). The system pulls news articles from APIs and converts numerical data into natural language narratives. Two summarization approaches were tested across three open-source models, with Falcon-7B showing the best results in accuracy and coherence. However, the study also highlights potential issues with repetition and hallucinated facts when using Retrieval-Augmented Generation (RAG). The system was built before RAG-based financial tools became widespread, but its findings remain relevant today. --- Why it matters: This research matters to engineers working on AI applications because it explores the limitations of Large Language Models in real-world tasks, such as summarizing complex financial data. Understanding these limitations is crucial for developing more robust and reliable AI systems. Source: https://arxiv.org/abs/2608.19526