Lomas: An Open-Source Platform for Confidential Analysis of Data (dgwk2026)

Lomas is an open-source platform developed by the Federal Statistical Office that resolves the tension between data privacy and transparency. It allows researchers and analysts to run algorithms on confidential public data without accessing raw information, using Differential Privacy to guarantee protection while maintaining full transparency. By keeping this solution open-source and publicly controlled, Lomas strengthens digital sovereignty and offers a concrete alternative to proprietary “black box” systems, demonstrating that democratic societies can have both privacy and transparency. Public administrations collect massive volumes of data to fulfill their missions, producing regional, national, and international statistics across many sectors. Yet strict privacy regulations, while necessary, often prevent broader analytical use of this valuable resource, creating a tension between transparency and privacy that democratic societies must navigate carefully. Lomas is a novel open-source platform developed by the Data Science Competence Center (DSCC) of the Federal Statistical Office (FSO) designed to resolve this tension by enabling the secure reuse of sensitive datasets for research and analysis. By developing it openly, we are creating a public good for the public good, strengthening both digital sovereignty and trust in the use of public data. The platform allows authorised users, such as approved researchers and government analysts, to run algorithms on confidential data without directly accessing the raw information. Results are returned with the strong privacy guarantees of Differential Privacy, a mathematical framework that quantifies and limits disclosure risk while maintaining complete transparency about data protection and usage. Through this approach, Lomas demonstrates that we don’t have to choose between privacy and transparency. We can have both. It offers a concrete alternative to proprietary “black box” solutions, keeping control over sensitive data within public institutions while making government data more accessible for democratic oversight and evidence-based policy-making. about this event: https://winterkongress.ch/2026/talks/lomas_an_open_source_platform_for_confidential_analysis_of_data/