🌍 Can we solve global issues without compromising data privacy?

This is what we discuss in our freshly published paper "Overcoming intergovernmental data sharing challenges with federated learning" at the Data & Policy journal.

Intergovernmental collaboration is key to addressing global problems, but data privacy concerns often hinder progress. Federated Learning (FL), a decentralized AI method, allows governments to use data from multiple sources without sharing raw data. Instead, only model parameters are shared, ensuring privacy.

In the paper, we discuss how FL can help overcome major data-sharing barriers identified by the OECD - OCDE. Plus, our practical example related to the Ukrainian refugee crisis demonstrates FL's potential for policymakers and researchers.

Cheers to the author team Kilian Sprenkamp, Joaquín Delgado Fernández, PhD, Sven Eckhardt and many thanks to the editors and reviewers.

The paper is here.