Ethical AI Principles for Enterprise Collaboration in Federated Learning Networks
Ethical artificial intelligence (AI) has received increasing attention as the development and use of AI applications have expanded. Collaborative machine learning, such as cross-silo federated learning (FL) can assist in adhering to ethical AI standards. FL ensures privacy and data sovereignty while minimizing model bias by aggregating AI models trained locally on data silos from several organizations. However, decentralization and multi-party involvement necessitate the refinement of existing ethical AI principles and the development or adaptation of methods for compliance. In this study, we conduct a systematic literature review followed by a workshop with participants from academia and industry on ethical principles for FL, considering technical and organizational aspects at different phases of the collaboration cycle. Our contribution is a guideline for technical and non-technical stakeholders to support the ethically aligned establishment/entry, value co-creation, operational continuity, and exit/termination of enterprise FL networks.
Researcher
Publication
, , , : Ethical AI Principles for Enterprise Collaboration in Federated Learning Networks European Conference on Information Systems (, 17. Juni 2024 – 19. Juni 2024)
URL: https://aisel.aisnet.org/ecis2024/track04_impactai/track04_impactai/16/
BibTeX: Download