|
|
Was this forwarded to you? Get The Fourth Estate AI Brief in your inbox →
|
|
|
|
View this issue in your browser
Tuesday Brief
Everyone's automating. Almost nobody's monetizing.
A new study of 160 AI projects at newsrooms found what you'd probably guess: we're using AI to transcribe faster and draft quicker, and almost nobody is testing whether it can help us make money. Meanwhile, Axios is placing a different bet, treating local reporting as data it can license to AI companies. Faster production saves time. Axios is betting licensing could save the business model. Worth watching which bet pays off.
|
|
THE AUTOMATION GAP
The Rundown: A new study tracked 160 AI projects at 101 news organizations. The pattern is clear: plenty of AI for production speed, very little for testing new ways to make money.
The details:
- Arizona State's Cronkite School analyzed AI projects from late 2022 through 2025 across 101 news organizations.
- Most adoption concentrated on transcription, summarization, drafting, and workflow automation.
- Business-model experiments (revenue, subscriptions, fundraising) made up only a small minority of projects.
- Every fundraising AI project came from newsrooms receiving American Journalism Project grants.
- Remove AJP grantees from the sample and business experiments nearly disappear.
Why it matters for us: If you're using AI for transcription and production, you're in good company. That's where everyone started. The study also points to an opening: nobody has cracked the revenue side yet, so a 10,000-circulation weekly willing to test AI-assisted subscription upsells or fundraising appeals has the field mostly to itself.
Read at tomorrowspublisher.today →
|
|
|
AXIOS'S DATA BET
The Rundown: CEO Jim VandeHei laid out a two-part plan: AI tools to support local reporters, and paid news feeds sold directly to AI companies.
The details:
- Axios is building an AI system specifically designed to support its local reporting staff.
- The company plans to sell verified reporting as paid data feeds to businesses and AI firms.
- Axios Local currently employs more than 100 journalists across its local coverage network.
- VandeHei argues AI will make human reporting more valuable as training data, not less.
- Additional revenue streams include executive-focused news and expanded in-person events.
Why it matters for us: For a community paper, this means your original council coverage and court records work might be worth licensing, instead of only being given away on Facebook. If Axios can sell verified local reporting to AI companies, can you? Or would you need to team up with your state press association to have enough volume?
Read at runtimewire.com →
|
|
|
WORDPRESS SECURITY CHECK
The Rundown: A speculative piece imagines AI speeding up both the discovery and the exploitation of website security flaws. That matters because most publisher sites run on PHP.
The details:
- The article walks through a scenario where AI finds basic PHP flaws that let attackers run their own code on your server.
- PHP powers roughly 70 percent of websites, including most WordPress installations.
- Recommended defenses: a web application firewall (WAF), automated vulnerability scans, and installing updates right away.
- Two-factor login on every admin account, and file permissions limited to what each user actually needs.
- Backups and a disaster-recovery plan are the last line of defense if someone does get in.
Why it matters for us: If your paper runs on WordPress (most do), take this list to your next IT conversation. Is a WAF in place? Have your automated backups actually been tested? Do admin accounts require two-factor? The AI angle is speculative. The hardening steps are real, and overdue at most small publishers.
Read at thetechedvocate.org →
|
|
|
TOOL EVALUATION 101
The Rundown: A sponsored guide to choosing AI sales tools has a framework that works just as well when your newsroom buys AI.
The details:
- Map your current workflow and bottlenecks before you evaluate any AI tool, not after.
- Run a pilot for at least 30 days with a small group of users before rolling it out to everyone.
- During the pilot, measure the time saved and how quickly the vendor answers support questions.
- Hidden data enrichment fees can push the total cost well past the first quote.
- The tool should sync both ways with the systems you already use, without lag.
Why it matters for us: Transcription, proofreading, ad sales automation: the same framework holds. Thirty days, a small group of users, the actual time saved, and how fast support responds. Most newsroom AI purchases skip the small test group and the time measurement, which is how you end up with expensive shelfware.
Read at oceanroadmagazine.com.au →
|
|
|
|
That's the week. If you're testing AI on the revenue side, I'd like to hear about it.
Trevor Slette runs Quadd.ai, which builds AI tools for publishers.
|
|
|
|
|
|
Run a newspaper. Use Quadd.
AI tools built for small newsrooms. Document intelligence that turns court reports and box scores into copy you can paste into your layout. Audio transcription that finds the quote without you scrubbing for it. AP-style proofreading that tightens copy without flattening voice. Built by someone who knows what a Tuesday night feels like. Free for seven days.
Request a 7-day trial →
|
|
|