The Tucson news you’d otherwise miss.
Tucson Daily Brief reads the agenda packets, watches the council meetings, and tracks the public filings — and brings you the Tucson stories hiding in the public record, every morning.
Tucson is one of the most under-covered metros in the country relative to its size. The Old Pueblo has more than a million people in its metro area and a handful of beat reporters covering everything from school boards to courts to development. More public business happens here in any given week than the existing outlets can possibly reach. Tucson Daily Brief exists as an experiment: to see how much of that gap one person, aided by modern technology, can fill.
What you’ll find here
The site is organized into a few clear sections. Here’s what each one is for.
Daily Briefs
Every morning, a synthesis of what the rest of Tucson’s press is reporting — city, county, courts, business, weather — gathered in one place, newest first. The fastest way to feel caught up before breakfast.
Local Government
What your local government is deciding — before and after. What to Watch reads every agenda packet ahead of a meeting and flags what’s worth watching; What They Decided covers what was actually decided afterward. Coverage spans the City of Tucson, Pima County, Marana and its planning commission, Oro Valley, the Rio Nuevo District board, and the Sahuarita Unified school board.
Around Town
What’s opening, building, and changing near you: new businesses, liquor filings, and food-establishment permits, plus rezonings and development cases — surfaced automatically from public records, most of which never get reported. Every item is tagged New business or Development. This is where our coverage of the fast-growing suburbs lives: development cases in Marana and Oro Valley, polled daily from each town’s own planning records, plus Marana’s liquor licenses straight from the state database, since the town never puts them on an agenda.
In Depth
Standalone feature stories on the issues that matter most across Southern Arizona, reported from TDB’s own archive of meetings, filings, and records, with the wider context layered in.
ChatTDB
Ask a question about Tucson and get an answer drawn from everything TDB has published — with citations. Think of it like ChatGPT, except it only knows, and only cites, TDB’s own reporting.
Tucson en Breve
The whole daily brief in Spanish at tucsonenbreve.com, plus a daily one-minute video newscast on TikTok and the day’s stories as Spanish cards on Instagram. Translated by a model, and it says so on every page.
The podcast
The Daily Brief is also a podcast, on Apple Podcasts and YouTube every morning.
How this is made
The newsroom is one person — me — which raises a fair question: how does one person cover six governments’ worth of public business before breakfast?
There is a thing in video games called a tool-assisted speedrun: a player uses software — frame-by-frame input, save states, automation — to play through a game with a level of precision and speed that no human could achieve in real time. The result isn’t laziness or cheating. It’s a demonstration of what the game makes possible once you stop pretending the constraints of human reflex are the constraints of the medium.
Tucson Daily Brief is that same idea, applied to local news. The tools are large language models, transcription engines, public-data scrapers, schedulers, and a few hundred lines of glue Python. Together, the run reads every agenda packet that drops, transcribes the council meetings as they happen, watches the liquor licenses, food permits, and development cases as they’re filed, and synthesizes what the rest of the Tucson press is reporting — most mornings, before breakfast. No newsroom is going to hire enough people to read all of it. So I wanted to see what a speedrun could do.
Here’s the run, concretely. Every morning at six, a model (Claude Opus 5, as of this writing) reads about two dozen local sources and writes the Daily Brief. Nobody reviews it before it publishes at 6:10, so every item links to the source it came from. When a council posts an agenda, a model reads the packet and publishes a What to Watch preview the same morning, also unreviewed, because it only summarizes what is already public. Around Town works the same way, and the newest businesses on it come from something simpler still: a daily check of the state liquor-license database and the county health department’s food-permit database, no model involved.
Everything that reports what actually happened is different. When a meeting streams, software transcribes it live and a model drafts a report from the transcript. That draft doesn’t publish itself. I read it against the transcript, edit it, and approve it, and the page names both the model that drafted it and the one that helped me edit. In Depth features are the reverse: I write them, and a model helps with research and drafting. The morning podcast is the brief, condensed by a model and read by an AI clone of my own voice, trained by me.
There’s video, too, and it’s Spanish-only by design. Tucson en Breve’s Roadrunner is a daily newscast of about a minute, built from the morning brief and read by Patricia, an AI anchor: a HeyGen avatar with an ElevenLabs voice, named for my adopted mother, who is from Colombia. Her script is translated by a model from the finished English script, never regenerated from the sources, and proper names are never translated. I listen to every episode before I post it to TikTok, each one labeled as AI-generated content.
Every article on this site ends with a note saying which model made it, from what, whether a person reviewed it, and when it published. Those notes aren’t typed. They’re generated from the pipeline itself, so they can’t drift from what really happened. The Spanish edition, Tucson en Breve, is translated by a model and says so on every page.
Whether any of this is “the future of news,” who knows. But it does show what one person can do when they treat AI not as a replacement for the work but as a toolkit.
You can also hear this experiment discussed out loud: in August 2026 the TWiT network had me on twice in one week — as a panelist on This Week in Tech, and as a guest on Intelligent Machines, where the chyron read “A One-Person AI Newsroom” — and KGUN 9 came out to Catalina to profile the project.
About the byline
My name is Nicholas De Leon. I’m a longtime technology reporter born and raised in NYC. I moved to Tucson in 2023 and quickly fell in love with the place! By day I’m a Senior Reporter at Consumer Reports and by night I work on TDB as well as a host of other interesting projects. The fastest way to contact me is by email. Tips, corrections, leads, and friendly hellos are all welcome.