This talk walks through my existential crisis engaging with AI. I left Google because I didn't believe nor respected the direction we were taking with Gemini, especially as a former fact-checker. I co-founded Indicator to cover digital deception and have repeatedly documented how AI has made it easier to harass people at scale and pump out sloppy engagement-bait. But over the past few months, I have found that Claude Code is a tool I can trust to have a transformative effect on my data collection, analysis, and representation. I'll walk through some (basic) examples of how my work has changed over the past year and lay out some open questions that my co-founder Craig Silverman and I are navigating as we build our AI policy.
I propose a lightning talk length case study of Bellingcat's work and our unique "audience engagement" strategy via an open-source investigative story about an oil spill in the Caribbean. This project started as a comment in our public Discord chatroom, and became a sprawling collaboration with multiple newsrooms across multiple continents. Along the way, it remained transparent and community-led. Our audience are more than passive consumers of our content. Through our public gathering spaces and volunteer programs, they can become co-creators of our research. By building this community, Bellingcat promotes ethical and responsible open source investigation skills, critical for media literacy and democratic participation in the misinformation age.
I will give quick take highlights on why this emerging field of creator priovders matters, who they are, their brackgrounds, business strategies, audience relationships, support structures and more. Data comes from long-form interviews with creators info providers in the U.S. and South Africa. The U.S. findings will be realeased April 13 with South Africa following soon after.
Executive Director, USA, Center for News, Technology and Innovation
Amy S. Mitchell is the founding Executive Director of the Center for News, Technology & Innovation. Prior to her role at CNTI, Mitchell served as Managing Director of news and Information Research at the Pew Research Center. In her 25 years with Pew Research, Mitchell helped launch... Read More →
Molly Stark Dean is a journalist and audience development leader who worked at CNN, Reuters, CBS, Fox News and CoinDesk. At Media Party NYC 2026, she pitches WikiNewsroom, a program that embeds Wikipedians in Residence in American newsrooms. This WikiProject would turn daily jour... Read More →
Traditional journalism is losing distribution. Creators are winning it. In this hands-on workshop, participants will transform journalistic content into creator-style formats optimized for reach, trust, and platform dynamics. Drawing on real-world experience from Dieter von Holtzbrinck Medien (publisher of Handelsblatt, Tagesspiegel, and Die Zeit), we will show how AI-powered workflows can turn high-quality journalism into scalable creator formats without losing credibility. Participants will analyze successful news influencers, extract repeatable patterns, and build their own format using AI tools (including secure enterprise AI and voice generation). By the end of the session, every participant will leave with a fully developed, ready-to-publish news format tailored to their audience and platform. This workshop is designed for journalists, creators, and editorial teams who want to compete in the creator economy without compromising journalistic standards. Workshop Structure (120 minutes) 0–30 min | Context, Framing, Pattern Analysis Why journalism is losing distributionAnatomy of successful news influencers (share insights of our whitepaper)Breakdown of real examplesPresent repeatable format patterns 30–60 min | Creation Sprint Participants define:audiencetopicformat conceptAI-supported ideation and structuring 60–105 min | Prototyping Vibe Coding 105–120 min | Share Results Gallery WalkPeer feedback and refinementClosing How is our session interactive?Participants actively build their own format throughout the sessionGroup work and peer feedback loopsAI-assisted co-creation with tech support (live)Real-time iteration and testingThis is a fully hands-on session, no passive listening. What will participants leave with?Each participant will leave with:A fully defined news creator formatA structured content blueprint (hook, format, distribution logic)A ready-to-use prototypeA repeatable workflow for future content creation
we detail where and how of potential product solutions to solve common challenges that news websites currently have (SEO, AI, bounce rate, reading time) using content/data APIs hosted by the Wikimedia Foundation
Most news organizations know they're losing young audiences. Few have real data on what actually brings them back. I do. Over the past year, Newsreel has been deployed in 200+ classrooms and libraries, with 41 student fellows running peer-to-peer news engagement programs across 25 universities including UNC, NYU, Northwestern, UC Berkeley, and Cambridge. Our month-one retention is 35%, which is 3-5x the news app industry average. A study of 277 Penn State students found 65% planned to keep using the platform after course requirements ended. In this workshop, I'll share the raw data behind what worked and what didn't, then participants will build their own audience strategy using the frameworks we developed. We'll work through three exercises: (1) mapping your current audience gap by analyzing who you're reaching vs. who you're missing, (2) designing an institutional distribution channel (schools, libraries, community orgs) that creates habitual use without paid acquisition, and (3) building a peer ambassador program modeled on our campus fellows playbook. This is not a panel about "reaching Gen Z." It is a working session where you leave with a concrete distribution plan built on real engagement data from real young readers.
