The paper presents a pilot study aimed at developing a novel approach for discovering, tracking over time, and summarizing the content of available publications on a given significant topic. A key result of this study is the proposed methodology for summarizing information from multiple articles on a specified topic by using large language models like BART and GPT with the help of fine tuning and prompt engineering. The developed approach has been applied in the creation of a software tool that collects publicly available information on a given topic from reliable sources for a specified period of time and displays in a concise form the most essential of the content of the publications found, thereby tracking the history of information on the topic.