Drs. Jason Millar (Canada Research Chair in the Ethical Engineering of Robotics and AI), Teresa Scassa (Canada Research Chair in Information Law and Policy), and Kelly Bronson (Canada Research Chair in Science and Society) are studying the ethical, social and legal aspects of COVID-19 contact-tracing applications.
It is clear that most successful public health measures aimed at minimizing the spread of COVID-19 have included rigorous contact-tracing protocols. That is, for each confirmed COVID-19 infection, public health workers conduct detailed interviews with each patient to determine precisely with whom that person has been in contact leading up to the positive COVID-19 diagnosis. Contacting and instructing those “contacts” to self-quarantine is an effective strategy for preventing further rapid spread of infection. Contact-tracing can also help support responsible strategies for “reopening society.” As people begin to socialize again—visiting restaurants and mingling in public places—vigilant COVID-19 testing and contact-tracing could help manage the baseline rate of infection until the disease no longer poses a public health threat.
To improve COVID-19 contact-tracing governments are considering the use of smartphone-based contact-tracing apps. These apps are currently being released;eased into the world, and use a combination of location-based services (GPS, Wi-Fi or Bluetooth) and information about users’ COVID-19 “risk profile” (e.g. infection status, prior contacts) to track a user’s movement and help determine if the user has come into close contact with “high-risk” individuals. When contact occurs, public health messages can then be sent to at-risk users in the system, including reminders to wash hands more vigilantly, or to self-quarantine.
Contact-tracing apps are as yet unproven, and they pose a number of significant social, ethical and legal challenges, each of which must be evaluated when designing the apps, and when balancing the potential benefits and harms of deploying them. There has been considerable concern expressed around privacy and contact tracing, including concerns over the ability for app users to identify who may have tested positive, the potential for reidentification of individuals from deidentified data, data security, and the use by governments of these apps to track individuals. Other concerns relate to the potential of an app to have a discriminatory impact on individuals or communities; the potential reproduction of harm to already vulnerable communities; widespread over-trust in apps leading to an increase in infection rates; individual psychological harms; notification burnout leading to less effective public health messaging; limited access among marginalized socio-economic groups (e.g. refugees and other non-English or French speakers) and indigenous peoples and communities; and overloading healthcare system resources if individuals respond unpredictably to app messaging.
These challenges should be carefully, and continuously, considered during the development and deployment stages to minimize resulting harms, and also to increase the overall trustworthiness, thus effectiveness, of contact-tracing apps. To be effective, AI-powered contact-tracing apps require significant uptake in society. One estimate suggests that 58% of a population must be using an app for it to be effective. Thus, failure to rigorously account for the potential harms associated with contact-tracing apps could foreseeably undermine their public health benefit.
Our team is working with teams developing these apps and policymakers to support the responsible and inclusive design and governance of COVID-19 contact-tracing apps.
We will update this page and the CRAiEDL website as our results begin to emerge.
This project is partly funded by the AI + Society Initiative at uOttawa’s Centre for Law, Technology, and Society.