At NEWSWELL, we’re working to transform and sustain local news in order to help communities and democracy thrive. Thankfully, other organizations also care about this type of work, including the American Journalism Project.
We were pleased to have AX Mina, AJP’s senior VP of specialized support, join us for the 2026 National Journalism + AI Accelerator — and recently asked Mina to share some best practices for AI use in newsrooms, what she’s working on now and what she thinks about the future.
Can you tell us about your role at the American Journalism Project?
I lead the Knight Resiliency Lab, a specialized team that aims to drive excellence and innovation across our grantees’ key functional areas by helping them grow and build long-term resilience.
Our Product & AI Studio, launched in 2023 to explore the smart application of AI and other technology within local news, sits within the Lab, which I think reflects something important about our philosophy: we see product thinking and working with emerging technologies as core competencies for nonprofit local news organizations today. It’s part of AJP’s broader venture support approach to strengthening local news.
Can you share some key learnings?
A few things have really stood out. First, this work has to be prioritized by leadership, in concert with their team, in order to be effectively applied. It also can’t be siloed in one department. The most productive conversations we’ve seen are multistakeholder ones, bringing together editorial, tech, revenue and operations, to determine effective guidelines and strategies for implementing AI — and also when to avoid it.
Second, we’ve noticed a real increase in confidence among newsrooms that engage seriously with AI, even if they start trepidatiously. The 29 organizations participating in our cohort programs spent time learning use cases, deploying responsibly and how to evaluate vendors, and they all indicated increased confidence over time.
Third, the use cases are varied. LLMs (large language models) can be used as a coach, kind of like a friendly guide to double-check work or surface best practices or as a tool baked into back-end editorial workflows, like doing a first pass on translation or fact-checking. And some organizations, like City Bureau and Open Campus, are actually designing an application layer on top of LLMs to make the work more impactful.
Any important lessons about newsrooms needing to be cautious with AI?
Transparency is essential, both internally with staff and externally with audiences. Organizations need clear guidelines and policies that are genuinely aligned with their values and developed with the right stakeholders at the table. We recently published a guide on this, building on great work spearheaded by our friends at Poynter.
I’d also encourage newsrooms to think of AI as both a tool and a topic. Their audiences are trying to understand AI too, and news organizations have a real opportunity — and responsibility, I’d argue — to help their audiences understand the impact of these tools in their daily lives. Every conversation about AI policy is also a chance to model thoughtful, values-driven thinking for the communities they serve.
What’s next for the Product & AI Studio?
We’re planning our next wave of cohorts designed to help advance innovation by newsrooms, which we’re really excited about. We’re also starting to build products ourselves that have potential to scale across our portfolio. And you can expect some new case studies and playbooks (down the line). We want what we’re learning to be useful to the broader field, not just the organizations we work with directly.
Anything else you’d like to share?
I think a lot about Amara’s Law, namely, that we tend to overestimate the immediate impact of a technology and underestimate its longer-term effects. And we’re still in the early stages with what we currently call “artificial intelligence.” The more we can step past that word and start talking about specific tools — LLMs, image generators, transcription engines, translation workflows — the better equipped we are to think strategically about when and how they make sense for a news organization — and whether they align with our values. These are technologies that require the usual product strategies: outlining pros and cons, determining clear use cases based on user needs and evaluating responsibly.
We appreciate Mina taking the time to share these insights, and we applaud the great work that’s happening at AJP.
As Mina noted, AI is still in its infancy — and now is the time for newsrooms to dig in, discuss, explore and experiment.
At NEWSWELL, we’re working to thwart AI-fueled disinformation in local news and partnering with our newsrooms to see how AI can help them streamline some of their processes.
This update appeared in our April 17, 2026, newsletter. This version has been lightly edited for clarity.