Balancing Trust and Safety on End-to-End Encrypted Platforms
While the idea of end-to-end encrypted communications has been around for decades, the deployment of default E2E encryption on billion-user scale platforms has major new impacts for user privacy and safety. Privacy against the prying eyes of governments and corporations used to be only available to those with the technical skills necessary to use esoteric software. Now, billions of people share the most private personal details in conversations that are protected from all but the most intrusive attempts by skilled adversaries, with no special skills or even intention to communicate privately necessary.
The deployment of default end-to-end encryption, most notably to Apple’s iMessage and Facebook’s WhatsApp, has come with major benefits to both individuals and society. It has also created new risks, as long-existing models of messenger abuse can now flourish in an environment where automated or human review cannot reach it. Recent announcements of new E2E encrypted products have raised the prospect of new, less understood risks coming to the forefront. For example, none of the large E2E encrypted networks currently allow user discovery via a person’s commonly used name, and allowing encrypted initial contact from strangers could increase the volume of certain types of abuse.
Through a series of workshops and policy papers, the Stanford Internet Observatory is facilitating open and productive dialogue on this divisive and controversial topic to find common ground and areas of compromise. An important defining principle behind this workshop series is the explicit absence of discussion of exceptional access (aka backdoor) designs. This debate has raged between industry, academic cryptographers and law enforcement for decades and little progress has been made. We focus instead on interventions that can be used to reduce the harm of E2E encrypted communication products that have been less widely explored or implemented.