Federated Learning: Organizational Opportunities, Challenges, and Adoption Strategies

  • 2023-09-06 13:21:22
  • Joaquin Delgado Fernandez, Martin Brennecke, Tom Barbereau, Alexander Rieger, Gilbert Fridgen
  • 0


Restrictive rules for data sharing in many industries have led to thedevelopment of federated learning. Federated learning is a machine-learningtechnique that allows distributed clients to train models collaborativelywithout the need to share their respective training data with others. In thispaper, we first explore the technical foundations of federated learning and itsorganizational opportunities. Second, we present a conceptual framework for theadoption of federated learning, mapping four types of organizations by theirartificial intelligence capabilities and limits to data sharing. We thendiscuss why exemplary organizations in different contexts - including publicauthorities, financial service providers, manufacturing companies, as well asresearch and development consortia - might consider different approaches tofederated learning. To conclude, we argue that federated learning presentsorganizational challenges with ample interdisciplinary opportunities forinformation systems researchers.


Quick Read (beta)

loading the full paper ...