Sponsored by
Hi, I’m media innovation journalist Ulrike Langer and you’re reading News Machines. If this newsletter was forwarded to you becoming a subscriber means you will never miss an issue. If you want to support my work please spread the word, become a paid subscriber or click on the ad.
Alexander Hall went from high school journalist to journalism tech founder. And he only just turned 21. Right: Views of the Related Articles and Metadata features in the Ahody dashboard.
Alexander Hall was 17 when he started working at Allt om Norrtälje, a free digital news site in a Swedish town of 65,000. At age 18 he became a 33 percent co-owner. His competition: Norrtelje Tidning, a 150-year-old Bonnier-owned newspaper. His team: two people. His advantage: none, until he built one. I talked to Hall via Zoom about the origins of Ahody and his plans.
Ahody started as a survival tool. Hall and his partner were manually refreshing around 100 different websites, arriving late on stories, letting articles fall through the cracks because someone forgot to follow up on an unanswered SMS. "We had the problem that you maybe send a question to someone and then they don't answer. And then you just forget about that article." The result was a stack of half-finished stories and a newsroom perpetually one step behind.
Today, Allt om Norrtälje publishes at least 10 articles a day and reaches roughly 50,000 monthly readers - nearly the entire population of the municipality. The majority of those articles are now produced with Ahody's help. Norrtelje Tidning is reportedly asking how they do it.
Sponsored
How Jennifer Aniston’s LolaVie brand grew sales 40% with CTV ads
The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.
How it works
Ahody is a pre-CMS editorial platform that automates sourcing, draft generation, context enrichment, metadata, image selection, and publication scheduling. It monitors sources the newsroom selects - in Allt om Norrtälje's case, around 40-50, including police alerts, court cases, municipal documents, company registrations, property sales, weather warnings, and animal sightings - and routes relevant items to a journalist inbox with an AI-drafted article attached.
What makes it more than a monitoring tool is the workflow that follows. Hall shared a concrete example: When Ahody picked up a report about a possible new primary care emergency service in Norrtälje, Ahody drafted an article and its context agent suggested automatically adding relevant background - identifying that Norrtälje had previously had such a service that was shut down years earlier. The platform then ran Ahody against the source material, a press-ethics review, language flagging for loaded terms, and proofreading, all within the same workflow. A human editor then reviewed the article before publication.
This is the media library of Allt om Norrtälje
The platform is also where the article lives through its full life cycle. Journalists can collaborate through a built-in chat attached directly to each article, run deep research from inside the editor, and manage work across stories on a canvas-based workdesk. Images are automatically tagged and alt-text is generated. Metadata populates automatically. When an article is ready, it publishes into the newsroom's own CMS with all fields intact — no copy-paste, no manual metadata entry. Newsrooms can also configure their own house style rules directly in the platform, grammar preferences, tone guidelines, terminology, which are applied automatically across all AI-generated drafts.
Two source types - company registrations and property sales - auto-publish without human review, with a clear disclosure label and a correction button visible to readers. For Allt om Norrtälje, everything else requires editorial sign-off. But publishers can decide for themselves where to do draw the line.
"It is not an AI-first platform," Hall says. "It is journalism powered by AI."
The feature that sets Ahody apart from generic monitoring tools is its auto-configuration agent. Point it at a new source - a government API, an agency feed, a public database - and it writes its own monitoring script. No developer required. You set it up once and the monitoring runs indefinitely. For smaller newsrooms without technical staff, this removes a barrier that has kept sophisticated source monitoring out of reach.
The sourcing agent is what sets Ahody apart from many other journalism AI tools
From internal tool to startup
The pivot from newsroom fix to product happened fast. A friend working on an AI automation project approached Allt om Norrtälje as a reference case. One week in, Hall saw the bigger picture. Ahody Labs AB was formally registered in November 2025 with a four-person founding team. Allt om Norrtälje and the local news website TelgeNytt are regular paying customers. Paid pilots are running with several Swedish publishers, including VK Media, others have expressed interest.
The funding story has a telling structure. Stefan Lundell, co-founder and editor of Breakit - Sweden's leading tech and startup publication and winner of Sweden's most prestigious journalism prize, Stora Journalistpriset, in 2021 - personally invested SEK 1.5 million (about 138,450 EUR or $153,630). Swedia Capital invested another SEK 1.5 million.
What makes Lundell's investment notable is the way it happened: Lundell got to see how Ahody works and was convinced. That pattern has a parallel worth watching. At Coastside News Group in California, two board directors personally invested in Locunity, the AI civic coverage tool the outlet uses and publishes. It is still rare and notable for news org founders and organizations to invest in tools that run their infrastructure. It may not stay rare and notable.
Hall told me that Ahody will become the main partner of the industry organization of Swedish magazines. Reaching international markets is on the roadmap for 2026, though the company has not disclosed specifics on how or where.
The dashboard for reviewing and managing articles
What's still ahead
At 21, Hall is the youngest founder News Machines has profiled. He openly acknowledges that the product depth comes from his CTO and CPO, guided by what the journalists tell them to build. That newsroom-first model is exactly what larger outlets like Aftonbladet have found when developing their own AI tools: The journalists must lead or the tools don't stick.
Hall himself names the right risk: A powerful auto-publishing platform in the wrong hands - without an editorial culture behind it - is a misinformation accelerant. Ahody vets its customers. That's a workable policy for a five-person company. It gets harder to maintain at scale, and Hall knows it.
Five take-aways for publishers
The sourcing bottleneck is the real constraint. Most AI journalism tools address writing. Ahody addresses what comes before writing - signal detection, source monitoring, context retrieval - which is where small newsrooms lose the most time. Publishers evaluating AI tools should ask whether the tool solves the upstream problem or just speeds up the downstream one.
Infrastructure decisions are becoming investment decisions. When a media company founder backs a tool or platform financially, the vendor relationship changes. Editorial dependence and financial interest become linked. Publishers need governance frameworks for this before it becomes the norm, not after.
Contextual memory is an underrated feature. The ability to automatically surface that a story connects to something that happened three years ago - without a journalist having to remember or search - is not a writing aid. It is an institutional knowledge layer. That is where AI adds value that scales with time.
Transparency architecture matters more than transparency statements. Ahody's approach - auto-published content clearly labeled, correction mechanism built in, human review required for everything else - is a structural design choice, not a policy promise. Publishers should evaluate AI tools on what transparency looks like at the workflow level, not just what the vendor says about it.
The competitive threat to established local news is not hypothetical. A two-person newsroom outpacing a 150-year-old regional daily on volume and speed is a concrete outcome, not a projection. Publishers who assume incumbent advantage will protect them from AI-native competitors are looking at the wrong variable.