The paper presents an attack against federated learning algorithms and shows that when certain conditions apply, it may be possible to reconstruct the raw data from the gradients. This is an interesting observation. Federated learning, despite not having any formal privacy guarantees, is gaining popularity in corporates that operate on large amounts of data. In some cases, it might be used under the assumption that it provides privacy. Therefore, showing that this feeling is wrong may have real world impact.

The numbers were mind-boggling to contemplate at this year's CogX AI Festival. By 2025 there will be 1036 (yes, that is 10 to the power of 36) IoT devices in the world that will be connected and sharing data. IoT will converge through AI, which will not be limited by time, only by computing power. Making decisions is where AI will have the biggest impact. 'This is the year AI moves from classifying the past to making decisions for the future' proclaimed Vishal Chatrath, CEO of Prowler, at the keynote.