Paper Number



Methods and software components for developing novel IT solutions based on artificial intelligence (AI) technology are broadly available to organizations of any size. However, as AI typically requires large amounts of data, smaller organizations are at a disadvantage compared to large competitors as the latter often have more training and test data at their disposal. Collaboration and data sharing between multiple smaller actors might offer a solution to this issue, but also poses a potential risk to privacy and confidentiality. Our research considers the concept of federated learning, which enables collaborative training without exchanging the actual data. Still, the benefits of value co-creation within federated AI ecosystems are unclear. To shed light on this topic, we present a data-driven analysis using the example of sales forecasting in retail. We show that three types of benefits can be expected in federated AI ecosystems, namely collaboration, privacy preservation, and generalizability.



When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.