Dataharvest 2026 - the European Investigative Journalism Conference

This is a hands-on workshop introducing the practical use of Earth Index, a tool that uses AI to search for user-defined patterns in satellite imagery across large areas, anywhere on the planet. It is a game-changer for environmental journalism, opening up major possibilities: from spotting hotspots of illegal mining, to quantifying industrial farming in a region, or mapping new roads pushing into forested areas.

During the session, we will do a guided walkthrough of the platform and work through step-by-step exercises to explore its core features: how to create a project, define an area of interest, generate positive and negative labels, run predictions, refine the results, and export findings. We will also share practical tips and best practices based on the Pulitzer Center’s methodology for successfully integrating Earth Index into an environmental investigation, including how to audit results, reduce false positives, and inform field reporting.

By the end of the workshop, participants will have a clear understanding of Earth Index’s strengths and limitations, and how to add it to their investigative toolbox.

IMPORTANT:

1. You must register in advance for this session to gain access to Earth Index. Please use this form to register:  https://docs.google.com/forms/d/e/1FAIpQLSe0CXmCcA3SZyVfDMHO3l9suLK_RwCJUqz_IU-oymMRAFVSuQ/viewform
2. Please bring a laptop with Google Earth Pro installed. You can download it here: https://maps.google.com/intl/es/earth/download/gep/agree.html

No previous experience or coding skills are required. Those who sign up will receive free access to the tool and will be able to keep it afterwards.

Do you know what the telling signs of "non-authentic behaviour" - bots - are on social media platforms? This practical, hands-on presentation is for journalists interested in investigating bots across platforms - from Telegram and YouTube to TikTok. We'll use data research on bot networks and coordinated activity to learn how to identify bots, and discuss practical stories one can develop based on the data found. We will also give practical examples stemming from our research in Ukraine and Moldova, and present some insights into bot campaigns ahead of the June 2026 election in Armenia.

A key part of the session will focus on how things are changing with the rise of generative AI and the early shift toward more agent-like systems: what this means for bots, coordination, and the limits of traditional detection methods. By the end of the workshop, participants will have a clearer sense of how large-scale CIB operations can look in practice, and how to start identifying and documenting them across different platforms. The goal is to leave the session with concrete story ideas, investigative ideas, and practical methodological knowledge.

No programming skills are required to take part in the workshop, but prior experience with collecting and analyzing data from social media platforms will help participants get the most out of the session.

“Fueling Ecocide” is a cross-border, collaborative data investigation coordinated by EIC (European Investigative Collaborations) and EIF (Environmental Investigative Forum). Over the course of a year, 13 media outlets across four continents joined forces, asking a simple but crucial question: how much protected land and sea have we lost globally to oil and gas extraction?

To answer this question, we embarked on a mapping exercise of epic proportions. Using QGIS and Post-QGIS, we compared 315,000 protected areas from all over the world with 15,000 extraction blocks, spanning 120 countries.

For the first time, this work provides a clear global picture of the damage: 7,021 protected areas in 99 countries overlap with oil and gas projects. This represents a surface of approximately 690,000 km², just over the size of France, most of it under internationally recognized protection statuses. Through extensive reporting, they also identified 763 oil and gas companies involved, with the largest contributors headquartered in Europe.

In this session, the reporters involved in the project will take you through the makings of a data-driven global investigation. They will talk you through the trial-and-error process of handling and harmonising an immense geospatial dataset: which tools they used, which methodological choices they made, and which obstacles proved the hardest to overcome. They will explain how the team turned this data into impactful stories, collaborated across the globe, and what lessons they learned.

 To read more on the project and the story, go here: https://ecocide.reportersunited.gr/

Non-consensual intimate content sharing is a growing problem, disproportionately affecting women and marginalized people, and increasingly amplified by deepfakes and automated tools. What looks like chaotic “leaks” online is often a structured, organized, and profitable ecosystem.

In this session, we’ll show how we uncovered networks that steal, share, and monetize image-based sexual abuse. You’ll see the technical approaches we use: tracking digital traces, mapping channels, analyzing hosting setups, entering abusive groups safely, and spotting patterns in offshore companies. We’ll also explain how to map networks of abusers by understanding their ecosystem of filehosters, cloud services, and CDN layers.

Finally, we’ll share practical tips to safeguard your well-being when investigating abusive material. By the end, participants will be able to turn overwhelming, chaotic content into actionable insights, exposing the organized systems that enable and profit from abuse.

Algorithms are everywhere, but they're still extremely opaque. They affect us more and more, and often have a meaningful impact on our daily lives—especially when the public administration uses them. But it's incredibly hard to 1) get data and information about how they work, 2) understand it, and 3) explain it in a simple, impactful way.

In this session, we'll share different techniques that Civio has used to research and report on algorithms. We'll look into the ways in which one can crack open the black box of algorithms, from FOIA requests to court battles, reverse engineering to scientific reports. We will also explain how we analyse their outputs and how we make reporting that is often technical and complex more palatable for our readers. In the presentation, we will mention some particular examples from our reporting on algorithms, such as exposing facial recognition in doctor offices, algorithm bias when it comes to cancer detection, and a silly lie detector used by police.

We'll also talk about how we went to the Supreme Court to get the code of a system which decides -wrongly- who receives subsidies, and who doesn't. This session is suitable for beginners - you don't need to have had any experience reporting on algorithms in order to attend and follow the session.