Privacy is a major concern is the modern world. AI solutions often use sensitive data, private conversations, personal images, health data and many more. Several big companies as Amazon, Google and Apple have already reported private data leaks, services were shut down and employees were fired. Another aspect of the problem is future development. Privacy rule, as GDPR, healthcare rule, companies’ secrets can impede the possibility to develop AI solution to various issues.

From user perspective, the problem also exists. Imagine any voice command device that can perform specific tasks like play a music or turn on the lights. Usually, all data is being sent to the cloud for analyzation. All our private data!! How do we do it differently? Can we avoid privacy violation as users and developers?

In this round table we plan to explain the different privacy aspects and discussed some solutions. Specifically, we will delve into MPC, Multi Party Computation that can compute a function without revealing the inputs. Only random data is being sent for analyses, but the random data is enough for the conclusion. We will also touch differential privacy need and solutions.