Privacy Optimization Meets Pandemic Tracking

The novel coronavirus pandemic has led to a global public health crisis that requires a response from the whole of society—smartphone software developers included.

Mobile apps have the potential to help trace and then slow the spread of the SARS‑CoV-2 (severe acute respiratory syndrome coronavirus 2) virus, which causes COVID-19 (coronavirus disease 2019). They also have the capacity to do so faster and while preserving more privacy than traditional contact tracing techniques—or they could end up as yet another episode of botched government procurement and application of technology.

Apple and Google’s system to track COVID-19 infections anonymously via Bluetooth low-energy beaconing between iOS and Android smartphones has pointed a spotlight on a needed debate about balancing privacy and collecting useful data. Prior attempts to track the spread of infectious diseases have relied heavily on both human labor and the construction of centralized databases, and more recent attempts to leverage smartphone location data and other digital traces of our daily habits still require the collection and centralized analysis of large amounts of personal data.

The Apple/Google API aims to leverage such privacy-optimizing techniques as federated learning and differential privacy, which would allow developers to build useful alternatives to centralized databases that may invite later abuse. But will public health agencies and their developers take proper ...