WeBank, Swiss Re in federated learning deal
Tencent-backed online lender WeBank and reinsurance giant Swiss Re’s Beijing branch have signed a Memorandum of Understanding on joint research on the application of federated learning in reinsurance.
The cooperation will focus on federated learning, one of the latest developments in the world of artificial intelligence. And the two parties will work together to explore how the technology can help address the challenges imposed by data silos, enhancing the further development of the insurance industry.
A data silo is where only one group in an organization can access a set of data, which will cause wasted resources, inhibited productivity and inefficiency.
Federated learning is an encrypted and distributed machine learning approach which enables training for joint machine learning on decentralized data, wherein no data transmission is required for participants.
This new approach builds training models that are in compliance with data security requirements enhancing the outcome of machine learning. As an emerging AI technology, it is expected to form the basis of the next generation of AI-powered collaboration networks.
WeBank said its AI team had created “Federated AI Technology Enabler,” the world’s first industrial-level open-source technical framework and pioneered the development of federated learning standards to promote interdisciplinary and intercompany cooperation, supporting the federated AI ecosystem across industries.
The Swiss reinsurance company said the partnership will lay a solid foundation for the cultivation of new businesses supported by federated learning, thus encouraging the industry to adopt and apply new frameworks to improve the technological innovation capacity of insurance solutions.
John Chen, president of Swiss Re China, believes the cooperation will greatly accelerate the development of data sharing and data utilization, and contribute to the upgrading of the insurance industry’s pricing model as well as the innovation of products and services.
