"Tool Clinics" - Embracing Multiple Perspectives in Privacy Research and Privacy-Sensitive Design

Computer scientists or engineers are continually asked to “solve problems” or “improve” existing situations, by selecting from available design features to produce the “best” technical solution. For example, a software developer faced with the problem of securing data must choose between different encryption algorithms – each with different characteristics. Factors such as strength of encryption, speed of encryption, usability, key management and hardware requirements must all be considered. Other requirements such as the sensitivity and amount of data to be protected, the estimated resources of potential attackers, the operational context of the required solution, etc. must also be taken into account. It is impossible for any solution to be 100% perfect, e.g. encrypting data with no detectable delay using an algorithm which cannot be broken. Trade-offs during the design and development process are therefore inevitable as requirements are balanced, e.g. speed vs. strength of encryption. These trade-offs are dilemmas faced by the specialist in arriving at the final design. However, what is the “best solution”, and who decides what “best” means, requires more involved discussion and reflection. The engineer, with their narrow focus on solving the technical problem, might not be best equipped to solely decide what the optimum solution is, particularly if there are likely to be unintended consequences when the solution is deployed, or the proposed technology is decoded differently by users, those directly or indirectly affected, and other stakeholders.