Local news starts becoming local infrastructure

In September 2006, Adrian Holovaty published a groundbreaking essay about the fundamental assumption of what local news should be — structured information “that can be sliced-and-diced, in an automated fashion, by computers” instead of “a big blob of text that has no chance of being repurposed.”

Holovaty imagined a world where a story about a local fire would let readers explore the raw facts: date, time, place, victims, fire station number, response time; then compare those details with previous fires, and subsequent fires, whenever they happen. The essay inspired PolitiFact, which won the Pulitzer Prize in 2009. But most newsrooms never made the shift.

Six years later, Jeff Jarvis wrote a frustrated post on his blog Buzz Machine. Hurricane Sandy had just devastated his New Jersey neighborhood. He needed to know which streets were passable, where power crews were actually working, which gas stations had fuel. His local news outlet published stories about the devastation but left the community on its own to find useful information to get through the day.

Holovaty described a vision for local news as community infrastructure. Jarvis showed the consequences of its absence. But for most local newsrooms, this kind of infrastructure required developers, custom databases, and budgets that most local news outlets didn’t have.

Fast-forward to the near future: AI has the potential to change the fundamentals of local news. In 2026, the news organizations that thrive will stop thinking of themselves as publishers and start operating as community information utilities. They won’t just report on the school board meeting. They will show you when it happens, let you search past decisions, and connect you with neighbors who care about the same issues.

Some pioneers are already doing this. Hearst’s DevHub built Meeting Monitor as an internal tool to help reporters track local government meetings they could not attend. The Houston Chronicle now offers it directly to readers. Chalkbeat monitors around 80 school districts in 30 states through its LocalLens account, turning AI transcriptions into story leads. Civic Sunlight started as a newsletter that summarizes local government livestreams for the general public, then expanded to serve small newsrooms.

Village Media in Canada takes this model further. “We no longer see ourselves as a pure media company, but as a ‘community impact organization,'” Richard Gingras, Google’s former vice president of news and Village Media’s board chair, told me. Their AI tools monitor local government bodies and filter information for relevance before forwarding it to editors. Their community platform lets residents follow topics the way Reddit does. Village Media now licenses its technology to more than 100 publishers.

This is becoming a pattern: AI helps build tools for services offered to citizens. The news outlet becomes a platform for strengthening the community.

I have spent the last year documenting how newsrooms implement AI for my newsletter News Machines. The most interesting projects share a common thread. They do not just make journalism faster or cheaper. They change what a news organization does for its community. Government meeting archives, searchable databases of local decisions: These functions once required dedicated developers and constant manual updates. AI lowers that cost dramatically.

Ken Doctor’s media brand Lookout in Santa Cruz and Eugene, Oregon uses AI to automate the assembly of hyperlocal neighborhood newsletters from public data sources like permits, roadwork, inspections, crime reports, weather, and events. But Lookout is more than a local news outlet. It is a platform for solving local problems together with the community, supported by local business partners. Lookout is planning to scale its successful model for local news into more markets in 2026.

German media executives I work with often ask what American local news can teach them. My answer has changed. A year ago, I pointed to subscription strategies and newsletter formats. Now I point to this infrastructure model. The question is no longer “what stories should we tell?” but “what services should we provide?”

The shift carries risk. Research from Trusting News shows that audiences react negatively when they learn AI was used in news production, even when humans remain in the loop. Building infrastructure while maintaining trust requires transparency about what AI does and what journalists still do.

But the alternative is worse. Local news deserts exist because covering certain areas costs more than the revenue it generates. AI can lower that cost and free up capacities to find new ways to engage with communities. Tools that once required a dedicated R&D lab now work through a Slack integration. A small team can monitor dozens of government meetings, maintain a community events calendar, and still produce original reporting.

The next time a hurricane hits, residents in communities with forward-thinking local news organizations will know which gas stations have fuel, which roads are open, and where power crews are actually working — not from articles, but from the AI-supported community infrastructure their local newsroom built. Twenty years after Holovaty described the vision, the tools to deliver it finally exist.