Dataharvest 2026 - the European Investigative Journalism Conference

In 2014, we decided to request a comprehensive dataset from the Swiss state: We asked for a detailed overview of all granted export licenses of arms and military goods. Over ten years later, we run the most comprehensive arms exports database in a European country based on the documents received: www.rüstungsreport.ch. In our workshop, we take participants on a journey of investigative research, legal struggles, methods, and outcomes.

We'll explain how we managed to create a database of all arms export licenses obtained for armaments and surveillance equipment from Swiss-based companies around the world. We'll take you through a lengthy access process via freedom of information requests. The information was obtained only after a long legal battle against the Swiss state, which we won before the highest Swiss court. We'll also show how we built a database, organised the data, and how this became a tool of transparency, as we kept requesting the data and updating the database every year. Finally, we will showcase stories that came out of the database, uncovering critical company deals and state practices, and their potential for cross-border stories. We'll guide the participants through our investigations into the use of PC-12 from the company Pilatus in the US war in Afghanistan. And how the toothless Swiss export controls in the dual-use sector, coupled with the strong lobbying of the arms industry, made it possible for the Russian war machine to rely on Swiss high-tech.

This session will enable the participants to adapt our methodology to their countries of interest and give them ideas for further investigations into the defence industry. We will hand out a reader with the most important learnings and a template of the freedom of information request we made.

Coding has long been a skill journalists wanted to learn to make their investigations more efficient and rigorous. The main barrier was the significant time investment required to develop that skill. But since large language models emerged, we no longer need to write code ourselves. We do, however, still need to make informed choices when instructing an LLM to write code for us. Otherwise, those choices get made for us by the model.

How do we instruct the LLM best? How can we understand a code? And how do we catch potential mistakes? No prior coding knowledge is required to attend this session. You'll learn a simple, systematic approach to conversations, context management, and effective prompting that will help you to code anything. The participants should have an account with a large language model provider (ChatGPT, Claude, Gemini or similar).

Journalists often work with sensitive information. This information should not end up in web-based tools like ChatGPT and similar services. However, there are alternatives: local LLMs that run on your own computer. This not only ensures data protection when processing large volumes of documents, but it can also save costs on expensive APIs.

This introductory workshop aims to answer the most important questions: What hardware do I need? What frameworks are available (LM Studio, Ollama, etc.)? Which models can I use for which tasks? And what does such a workflow look like (e.g., with Python)? This session is a mix of presentation and hands-on elements.

To attend this session, no prior knowledge is required. If you want to participate in the hands-on parts, make sure to download and install Ollama and/or LM Studio and download a local model like Qwen3.5-4B

After attending this session, the participants will understand the pros and cons of using local AI models and get ideas from real-life examples on how to use this knowledge.