CT Mirror AI for local journalism — Poynter case study

The Connecticut Mirror, a nonprofit news organization, hired an AI data reporter and product developer to build tools that extend the newsroom’s reach across 169 towns that its beat reporters cannot physically attend. The role, funded by the American Journalism Project, represents a growing model for how small newsrooms can use AI to close coverage gaps.

Context

Angela Eichhorst’s work at CT Mirror focuses on generating leads and extracting material from municipal meetings using AI transcription and summarization tools. The newsroom already had four internal AI tools built by its data editor, trained on different Connecticut-specific datasets: state laws, the state’s Blue Book, and other public records. These tools answer questions about state governance and surface information buried in large document sets.

The newsroom is also developing a large-scale video scraping tool that will transcribe and summarize video from government websites, allowing reporters to monitor meetings they cannot attend in person. The approach treats AI not as a replacement for reporting but as a way to find stories that would otherwise go unnoticed because of resource constraints.

Key takeaway

For small newsrooms, AI’s most practical application is not writing stories but finding them. By monitoring municipal meetings, scanning documents, and surfacing leads that would otherwise require staff time the newsroom does not have, AI extends coverage without replacing the reporting and writing that make the coverage valuable.

Who should read this

Editors and publishers at local or nonprofit news organizations exploring how AI can address coverage gaps, and writers interested in how AI changes the relationship between reporting capacity and coverage quality.