By Aldana Vales, AI J Labs Program Lead
“Participants will work on and present a final project focused on an AI topic of interest. Do you already have an idea about what topic you’d like to dig into?”
In September 2025, we asked this question in the application form to the AI Journalism Lab: Leaders. Almost 250 applicants had their own ideas. A chatbot for election coverage. A scraper for government leaks. A fact-checking assistant. We read every single response and tagged each one by where the applicant was placing artificial intelligence: Was the technology accelerating something their newsroom already does, enabling something previously impossible, helping teams align or something else?
Two thirds of the applications clustered around three buckets: efficiency, story discovery/newsgathering and audience engagement. These were tool-shaped problems: “I want AI to do X faster.” “I want AI to help me find Y.”
We accepted 23 of those applicants into our 2026 cohort. Four months later, now that the strategy, upskilling and scenario planning classes are behind us, we see that true understanding of how to make sense of artificial intelligence goes well beyond building a tool.
“Leadership has always been about making better decisions under uncertainty. What AI changes is the scale of that uncertainty — and the urgency of creating clarity for your team,” says Anita Zielina, lead instructor at the AI Journalism Lab: Leaders.
It’s no surprise then that 22 of the 23 final projects ended up pointing inward, at the organization, the team or the leader herself. Using the same framework we used on the applications, we saw that the workflow categories collapsed from 64% of applications to 17% of final projects.
Fourteen of the 23 projects, instead converged on some version of the same pain: a newsroom that was not ready, not aligned or did not know how to use this technology. This meant that leaders had heard about AI and seen the demos, but had not figured out yet how to move their organization toward using it well. That gap is what most of the Leaders participants chose to spend three months on.
Very often, the type of problem determined the type of solution. Every single alignment project in the cohort, for instance, produced a working tool to bring teams closer. Elisabeth Gamperl at dpa prototyped an AI-powered stakeholder management assistant for new leaders trying to earn trust.
“Stakeholder management is one of the highest-leverage skills a leader has — and it’s also one of the hardest to do well under pressure. Every conversation has two layers: the content of your project and the dynamics of the room. Getting both right, consistently, across many different conversations and context switches, is where most new leaders lose time and momentum. This assistant acts as a thinking partner that helps you prepare fast, stay sharp, and walk into every meeting ready to create real alignment — and to do your best, most impactful work from day one,” Gamperl explains.
When the problem was enablement, leaders took a different path. They created training plans, working groups or structured conversations.
“The gap at InvestigateWest wasn’t a lack of AI tools — it was a lack of shared norms and trust. That was my key challenge as a leader. We had no policy or guardrails, usage was uneven and sometimes quiet, and a lot of staff felt anxious or resistant because the stakes around accuracy, credibility and editorial judgment are so high. In a small, fully remote newsroom, focusing on people and process felt like the only responsible way to bring everyone along — and to make any eventual pilot or tool actually stick,” says Jacob Fries, executive director at InvestigateWest.
When the problem was strategic readiness, leaders wrote roadmaps and playbooks. Foundational documents for an organization. That’s exactly how Juan Carlos Simo approached this work: “At La Nación, we believe the real transformation AI enables is cultural and infrastructural, not technological. We have our own tools, ongoing experimentation and training, and we are at a point where we can think and build from a different stage. A shared playbook will give us the clarity and foundation to create not just faster or more of what we are used to, but what we never thought possible”.
There is, of course, room for improving workflows and efficiency. These projects hint at where the work goes once the internal, cultural scaffolding is in place. Aisha Farrukh, Head of Content at The Centrum Media, created custom GPTs capable of generating grant proposals based on her organization’s proven strategy.
“For the TCM team, this AI integration unlocks a level of operational efficiency and scalability that was previously challenging. By deploying custom chatbots trained on identical data and prompt structures, the team can dramatically reduce proposal turnaround times while enforcing strict consistency and quality across all outputs. More importantly, it unlocks a framework for multi-sector growth,” Farruk explains.
Beyond a tool or a playbook, leadership in the age of AI looks like someone who can listen to their team, name the actual problem and pick the right shape for the answer. As Zielina reflects, “The leaders who got the most out of this program did not leave with a better chatbot. They left knowing exactly where their organization was stuck and what it would take to move it.”
You can learn more about the projects by watching the public showcase of the AI Journalism Lab: Leaders below, or join our mailing list to receive more information and updates about the AI Journalism Labs.