Abstract

The availability of unprecedented amounts of digital data, combined with increasing awareness by citizens and companies about the value of their data, call statistical institutions to adopt stronger mechanisms to ensure privacy and transparency of the statistical production process in order to preserve public acceptance and trust. Some novel technologies that have emerged recently at the intersection between the fields of cryptography, distributed systems and computer science, are instrumental to achieve this goal. The family of so-called Privacy Enhancing Technologies (e.g. Secure Multiparty Computation, Homomorphic Encryption and Trusted Execution Environment) open new possibilities in terms of how “data” and “computation” are handled and controlled within and across organisations. With these technologies, governance policies can be enforced technologically without necessarily relying on a single point of trust. These technologies have reached maturity and are already moving from laboratories to real-world applications. An increasing number of statistical institutions have started to pioneer possible applications of these technologies in explorative or pilot projects around different use-cases. The goal of this session is to provide a stage for such pioneering activities, share early experiences and lessons learned, discuss what we can (or can not) expect from these technologies and exchange views as to how the adoption of these technologies impacts (or is impacted by) organisational aspects, business processes and legislation. The session will consist of 4 speakers reporting on experimental and pilot projects leveraging PET for statistics, including one speaker from the UN PET Lab. The session will be divided in two parts, with a series of brief presentations by the speakers followed by a panel discussion between the speakers.