Susan Catron wasn’t looking to experiment with AI. As managing editor of The Current, a nonprofit investigative newsroom covering Georgia’s coastal communities, she had watched general-purpose tools like ChatGPT produce plausible-sounding nonsense often enough to know the risks.
What do 1,000 journalists and PR pros know about AI that you don't? They took AI Quick Start, a 1-hour live class from The Media Copilot. 94% satisfaction. Find out how to work smarter with AI in just 60 minutes. Get 20% off with the code AIPRO: https://mediacopilot.ai/
Key Takeaways
- Susan Catron banned AI at The Current, then adopted Nota after vetting.
- The 10-person nonprofit started with one feature and scaled with trust.
- Skeptical small newsrooms can adopt AI by picking journalism-specific tools.
“I was the one screaming across the newsroom: ‘Don’t touch AI, leave it alone,'” Catron says.
But her 10-person team faced an impossible calculus. Founded in 2020 to fill the vacuum left by consolidated and shuttered local newspapers, The Current aimed to hold public servants accountable through in-depth reporting. That mission required time—time that was instead consumed by the repetitive mechanics of digital publishing. Reporters spent hours each week crafting SEO-optimized headlines, generating newsletter summaries and managing metadata. Unlike larger outlets with dedicated digital teams, The Current‘s journalists handled everything themselves.
Andrew Haeg, network product manager at the Institute for Nonprofit News, describes this as the “fried and frozen” problem plaguing small newsrooms: Journalists are fried from running at maximum capacity and frozen by the fear of wasting scarce resources on technology that doesn’t deliver. For Catron, the question wasn’t whether AI could help—it was whether any AI tool could be trusted with journalism. That search led her to Nota, an AI platform built by journalists specifically for newsroom workflows, trained on journalism data rather than the open internet. “That was my selling point with Nota,” she says. “I don’t know what the other ones do. I just know I trust this one right now.”
Starting with skepticism and a single feature
Catron’s approach to implementation reflected her newsroom’s caution. Rather than rolling out Nota’s full suite of AI tools at once, she began with headline optimization—a single feature that would let her team evaluate the system’s understanding of journalism without risking editorial integrity.
Nota’s headline tool offers three suggestions that editors can accept, revise or ignore. Over time, the system learns from these editorial choices, tailoring future suggestions to match the newsroom’s voice and preferences. This design aligned with Catron’s requirement for human oversight at every step.
The setup took less than an hour. Nota integrates directly into WordPress and other content management systems, eliminating the software installation and training overhead that often derail technology adoption in resource-strapped newsrooms. Catron uploaded 10-15 representative articles to establish The Current’s tone, then began testing suggestions against her own editorial judgment.
“It’s only as good as what we put in it,” she says. Well-reported and well-written articles yielded better AI outputs, reinforcing rather than replacing the newsroom’s commitment to quality journalism. Skilled editors remained essential for catching the small errors that occasionally surfaced.
Expanding to SEO tasks once trust was established
After several weeks of testing headlines, Catron began adding Nota’s other publishing tools to The Current‘s workflow. The platform generates SEO-optimized tags, slugs and meta descriptions—tasks that previously required editorial attention but rarely benefited from editorial expertise.
Unlike general-purpose AI tools that might hallucinate facts or inject marketing language, Nota works exclusively from content journalists have already written and fact-checked. The system doesn’t generate original copy. Instead, it reformats and optimizes existing work for different distribution channels.
This distinction mattered to Catron’s team. Nota’s underlying model is trained on journalism-specific data from sources the company has explicit rights to use. Josh Brandau, Nota’s CEO and former chief marketing officer at the Los Angeles Times, says this targeted training produces hallucination rates “way less than 10 percent”—a crucial improvement over general-purpose tools.
The journalism focus shows in features designed specifically for newsroom needs: C2PA content tagging for transparency, platform-specific formatting for social media and integration with newsletter tools. Each feature addresses a task that consumes time without requiring deep editorial judgment—exactly the kind of work Catron wanted to automate.
Building social media capacity without hiring staff
The Current began using Nota’s social media tools for approximately half of its posts. The platform generates platform-specific captions for X, Instagram and Facebook, adapting tone and length to match each network’s conventions while maintaining the newsroom’s voice.
For a small team juggling reporting, editing and community engagement, this represented significant capacity expansion. Social media presence matters for audience development and reader trust, but crafting effective posts requires time The Current couldn’t consistently spare.
Nota’s social tools work from published articles, generating multiple caption options that editors review and approve. Like the headline feature, the system learns from editorial choices over time, reducing the revision needed for future suggestions.
Catron sees additional potential in using Nota for community calendar updates—a feature that could both serve readers and open new revenue streams, but one her newsroom currently lacks capacity to maintain. The AI-assisted approach could make previously impossible projects feasible.
Establishing editorial guardrails and transparency policies
As The Current expanded its use of Nota, Catron established clear policies about AI use and transparency. The newsroom made AI usage policies explicit early in the implementation, recognizing that audiences increasingly expect disclosure even when trust benefits remain unclear.
Research shows that while AI disclosure alone may not increase audience trust, the absence of disclosure damages credibility. The Current’s approach pairs transparency with consistent human oversight—every AI-generated headline, excerpt and social post receives editorial review before publication.
This human-in-the-loop approach addresses both quality control and ethical concerns. Nota operates on a closed-loop system that doesn’t train on newsroom content without explicit consent. For publications with strict privacy commitments to sources and subjects, this data handling stands apart from general-purpose AI tools that may use submitted content for model training.
The platform employs enterprise-grade security aligned with SOC 2 Type II standards, including data encryption in transit and at rest, secure authentication with single sign-on support and role-based access controls. Transparency features like usage reports and granular access controls help newsrooms maintain oversight of how their content is handled.
What didn’t work—and how they adapted
The documentation doesn’t specify particular implementation challenges The Current encountered beyond Catron’s initial skepticism about AI tools in general. The newsroom’s measured, feature-by-feature rollout appears to have prevented the adoption friction that often undermines technology implementations.
However, Brandau acknowledges broader challenges facing newsrooms considering AI adoption: “It’s a process. To see the full value you need to use it at scale, consistently.” Small newsrooms must balance the upfront time investment in training AI systems against immediate publishing pressures—the same resource constraints that make efficiency tools necessary in the first place.
The results
The Current now uses Nota to handle most SEO tasks that previously consumed hours of staff bandwidth each week—generating headlines, excerpts, tags and slugs. The team uses Nota’s social media suggestions for approximately half of their posts.
“I don’t even remember how much we spend on it a month, but I’m sure it has saved me that much time,” Catron says.
Specific time savings or productivity metrics beyond this qualitative assessment are not documented in available materials.
What’s next for The Current
Catron plans to integrate AI literacy into The Current‘s summer internship program, teaching the next generation of journalists both how to use AI tools effectively and how to maintain healthy skepticism about their limitations.
“I’m going to put the fear of God into them about a few things,” she says with a laugh. “AI can allow us to get better and grow more trust. But it could also kill that trust in milliseconds.”
At the network level, INN has leveraged Nota to expand member content reach across its 500-plus nonprofit newsrooms. INN newsrooms generate approximately 26,000 stories monthly, many under open licenses for republication, but insufficient resources often limit wider distribution. INN uses Nota to create automated workflows that summarize member stories into editorially vetted excerpts for services like Text Rural—an SMS news service for rural communities—and On the Ground, a weekly digest of local political stories. The entire process requires just 15-30 minutes weekly for review and approval.
“You could effectively support a really good product for your audience in minimal time,” Haeg says. “That just wouldn’t be possible without AI—in this case, Nota.”
Small newsrooms considering similar implementations can explore Nota’s grant-backed pricing at heynota.com. Qualifying outlets with fewer than seven full-time employees and annual revenue under $250,000 can access the platform for $99 monthly.
