Axios built a custom GPT to run local news in one-reporter cities

The Axiomizer and the component-by-component strategy

The most concrete artifact in this conversation is the Axiomizer, a custom GPT where reporters drop drafts and get sharper headlines plus tighter versions of Axios's signature sections — 'Why it matters,' 'What's next,' and 'Between the lines.' Editing and style checks are being folded into the same tool so copy editors spend their time on judgment calls rather than formatting fixes.

What's worth noting is the deliberate scope. Murphy says they broke the newsletter into components rather than trying to automate the whole thing at once, and that 'the more specific the task, the better the results.' That is an engineering stance, not a marketing one: narrow, well-defined tasks are where current models are reliable, and Axios is building its workflow around that constraint instead of pretending it doesn't exist.

The news roundup is the clearest example. Rather than asking a model to guess what matters locally, Axios captured each reporter's actual process — which neighborhood blogs, regional outlets, and niche sources they trust — and encoded that into prompts. The output is a short vetted list the reporter still curates. The human judgment moved into the prompt; the drudgery of assembling candidates got automated.

The one-reporter city is the actual product

The strategic center of the announcement is a business-model claim, not a technology one. Axios says it can now launch a city with a single reporter and no extra production layer, and it has already done so in Boulder and Huntsville, Alabama — described as its first one-reporter cities.

Murphy is blunt that the local news crisis is an economic one: journalism has to be tailored to each community, which resists the copy-and-paste cost efficiencies other industries rely on. The pitch is that AI changes that math by removing costs that don't add value for readers, letting the newsroom reach places that couldn't previously support a bureau.

If we can launch a new city with just one amazing reporter—without needing a whole extra layer of production and support—we can go to places we never could before.Montana Labs

That reframes the whole exercise. The AI tooling isn't the deliverable; the deliverable is a lower per-city cost floor. The Axiomizer and the roundup prompts exist to make one reporter viable where two or three were required before.

Turning public-but-inaccessible records into leads

A second use case is more investigative in flavor: summarizing city council meetings, school board recordings, and government transcripts that are technically public but practically unwatched. Murphy frames this as spotting where a story is moving so a reporter knows who to call, rather than sitting through a three-hour meeting.

There's a parallel data-tooling effort — clean charts, 'vetted math,' and transparent comparisons on housing prices and school performance — meant to standardize the analytical layer while keeping the reporting local. The distinction Murphy draws is precise: standardize everything around the reporter, never the reporter's voice.

The reader-feedback loop follows the same logic. Axios runs quarterly surveys across all its cities but has only one audience insights lead. Analysis that used to take weeks now produces a one-page summary per city in under a day, which is a bottleneck-removal story more than a content-generation one.

OpenAI is funding the experiment it's being used to run

The detail that reframes this as more than a customer testimonial sits at the bottom of the piece: OpenAI has partnered with Axios to fund the expansion of Axios Local to cities including Pittsburgh, Kansas City, Boulder, and Huntsville. The vendor is financing the rollout of the workflow that showcases its product.

That alignment cuts both ways for applied teams reading this. The favorable read is that Axios's decomposition approach — encode expert judgment into narrow prompts, automate the busywork, keep the human at the center — is a repeatable pattern that would work with or without the funding. The cautious read is that the economic viability of the one-reporter city hasn't yet been proven on Axios's own dollar. When the tooling and the growth capital come from the same source, the sustainability claim is still an open question, not a settled result.