Can You Engineer Privacy?

Industrial-size data spills, leaks about large-scale secret surveillance programs, and personal tragedies due to inappropriate flows of information are guaranteed to have at least one consequence: engineers will be increasingly expected to integrate "privacy solutions" into the systems they are building or maintaining.⁴ Yet the task of engineering systems to address privacy concerns can be bewilderingly complex.

The seeming unwieldiness of the engineering task becomes evident in the concept of privacy itself and how this concept is negotiated. As a legal concept, privacy is defined rather vaguely. That vagueness, some argue, is part of its protective function. The open-ended definition allows people to invoke privacy as a category to protect their personal lives and autonomy from intrusions by others—including the state that endows them with citizenship rights and runs surveillance programs. European Data Protection Directive (DPD) or Fair Information Practice Principles (FIPPs) on the other hand are procedural measures, such as notice and choice, data retention limitation, and subject access rights. These principles are seen to be instrumental to making the collection and processing activities of organizations transparent. Although less ambiguous, data protection principles still need to be translated into technical requirements and are vulnerable to narrow interpretations. Moreover, FIPPs fall short of mitigating all the privacy concerns of users toward a given organization. They also do not address privacy concerns users may have with respect to other users, with people in their social environments, and toward a greater public.