Dataharvest 2026 - the European Investigative Journalism Conference

Research shows that, among audiences in Western countries, news avoidance has been growing nonstop for the last decade. The result? Sometimes, it feels that the main audience of our investigations is... other investigative journalists. While experts do use our research, it seems that we are failing to reach generalist audiences and the wider public.
One possibility to reach more diverse audiences is to present and disseminate our investigations in unconventional ways.

In this session, we will talk about using creative formats to publish investigative journalism: from unusual digital publications to printed and material objects, and live and in-person activities.

Come to this session to discover different creative formats and real examples of investigations that used them, as well as the lessons learnt. We will also present an online catalogue of creative formats, and show how editors and journalists can use it in their work.

More and more OSINT and data-driven investigations are being turned into documentaries for television and platforms. From the success of programs such as Cash Investigation in France, to Sweden’s flagship investigative show Uppdrag granskning, and internationally acclaimed documentaries such as the Oscar-winning Navalny (based on Bellingcat’s data-driven investigation) and the Emmy-winning Pegasus, data- , OSINT, and leak-based reporting is increasingly finding its way onto the screen.

But many journalists (specifically those who are not familiar with video production) are unsure whether their story is suitable for a documentary, or how to approach pitching, filming, security, or narration. This panel offers a practical methodology for turning an investigation into a documentary for journalists coming from print or online media. Drawing on concrete European case studies, we will explore how to pitch a documentary project to producers and broadcasters, when to start filming, how to protect sensitive data, and how to transform abstract data into compelling visual and narrative-driven storytelling.

After this session, participants will be able to:

-Assess whether a data/OSINT/leak-driven story is suitable for documentary storytelling
-Decide when and how to start filming during a data investigation
-Integrate audiovisual constraints (security, budget, narration) into their investigative workflow
 -Understand how to pitch a documentary based on a data investigation

When covering a long-running conflict, finding new ways to tell the story and keeping audiences engaged is a big challenge, especially when the stakes are too high for events to pass unnoticed. As Russian journalists in exile, we have been reporting on Russia's aggression against Ukraine for more than four years. For the third anniversary of mobilisation, which sent hundreds of thousands of Russian men to war, we developed a new approach to illustrating the consequences of that aggression for Russian society itself. By narrowing our focus to one of Russia's 85 regions and drawing on a range of sources, we were able to establish the most precise casualty figures ever reported at that level, combining that data with carefully crafted storytelling. The result proved effective: the piece reached not only Russian audiences, but widely read globally as well.

Attendees will leave this session with both the inspiration to keep reporting on long-running events even when audiences show signs of fatigue, and concrete techniques for doing so. These include thinking outside the box, revisiting older sources that may hold new value, and combining different reporting methods -data analysis, interviews, and OSINT- to keep coverage of a long conflict fresh, rigorous, and compelling.

Every slide in this presentation is hand-drawn. So is the thinking behind Cheatsheet, The New York Times' AI tool built from scratch for investigative reporters. We'll show what we built, in which cases we ditched off-the-shelf AI, and how we tested it with non-technical journalists.

The presentation will offer a practical roadmap for use cases (needle-in-a-haystack document search, mass translation, statistical analysis) plus an honest Dos and Don'ts list for AI in journalism and a first look at our open-source plans.