Events

When decentralization, security, and privacy are not friends

Speaker: Profs. Carmela Troncoso (EPFL)

Location: 60 Fifth Avenue, Room 446
Videoconference link: https://nyu.zoom.us/j/94706733828

Date: Wednesday, December 18, 2024

Decentralization is often seen as a main tool to achieve security and privacy. Decentralizing has brought these properties to anonymous communication systems, contact tracing applications, cryptocurrencies, and many others. Thus, it is not a surprise that a new trend of machine learning algorithms opt for decentralization to increase data privacy. In this talk, we analyze decentralized machine learning proposals and show how they not only don't improve privacy or robustness, but also increase the surface of attack resulting in less protection than federated alternatives.