2026 Reports

Mapping Charlotte: Using AI and an expanded definition of local news to better understand the civic information landscape

Stonbely, Sarah; Mehta, Dhrumil; Krisch, Michael; Aubin Le Quéré, Marianne

When local news mapping gained prominence a decade ago, the widespread concern was with news deserts. Now that local news—from journalism outlets, civic information providers, or independent newsfluencers—is available via social media, newsletters, YouTube, and elsewhere, people are more likely to feel adrift in information oceans. This paradigm shift requires us to reinvent our understanding of local news provision.

At the same time, the advent of artificial intelligence and advanced computational tools has allowed local news mapping to make giant leaps forward. Research can now go beyond the traditional local news outlet census to analyze content for both breadth of topics covered and detailed geographic coverage areas—down to the census-tract level.

In addition, it has become clear that people no longer rely on local journalism outlets alone (where they exist) for important local news and information. Accordingly, we broaden the scope of “who counts” as a local civic information provider to include civil society organizations—schools, houses of worship, municipalities, nonprofits—that are vital and trusted interlocutors at the local level.

This report provides a proof of concept, using Charlotte, North Carolina, as a test case, for both AI-driven methods for mapping news and the paradigm shift toward greater inclusivity. We found that Charlotte has myriad local news and information providers that span topics, neighborhoods, and platforms, and we used our newly developed tools to gain visibility into how specific segments of local news providers add value, and where conditions can be improved. We conclude with recommendations for Charlotte’s local news funders and publishers and for improvements to our research program going forward.

Topline findings

Charlotte is served by more than sixty local news providers, on a range of media and platforms, that cover a breadth of local-interest topics. Several niche providers cover topics not covered by anyone else in the local ecosystem. We visualize this using our semantic map.

The amount of coverage varies significantly by neighborhood, with Center City and other areas that are home to large event venues, hospitals, and the airport receiving a disproportionate amount of coverage while many other parts of the city receive very little.

When the least-covered neighborhoods are mentioned in local news, it is often in the context of crime stories.

Related, we found that crime and entertainment coverage vary predictably with the proportion of Black and white residents; neighborhoods with a higher proportion of Black residents are more likely to be covered in the context of crime, while neighborhoods with a higher proportion of white residents are more likely to be covered in the context of entertainment stories.

We analyzed the digital communications of more than five hundred Charlotte-serving civil society organizations (CSOs) to understand whether they are meeting local information needs in a way that supplements or fills in for missing local news. We found that CSOs are covering some topics—especially in schools and public institutions, such as firehouses—that local news outlets are not, suggesting opportunities for collaboration or content sharing.

Most Charlotte CSOs are very active on social media, with more than 80 percent of our sample using three or more platforms to reach their constituents. The most common platforms are Facebook and Instagram; 60 percent have email newsletters.

Overall, our mix of organizations was top-heavy with local news outlets, which limited the extent to which we could compare CSO coverage with local news. As a result, we learned valuable lessons that can be applied to future iterations of this method.