News: See our paper about our recent findings here: Decentralized is not risk-free: Understanding public perceptions of privacy-utility trade-offs in COVID-19 contact-tracing apps and summary of our findings in a tweet:

You probably have heard of many updates about #COVID19 #ContactTracingApp recently, including the APIs jointly developed by @Google and @Apple, and many proposals such as #DP3T, BlueTrace, etc.
— Tianshi Li (@tianshi_li) May 26, 2020

We are a team of privacy and human-computer interaction researchers working on a survey study to help related stakeholders build COVID-19 contact tracing apps by understanding the preferences of the general public.

Contact tracing apps have potential benefits in helping health authorities to act swiftly to halt the spread of COVID-19. However their effectiveness is heavily dependent on installation rate, which may be impeded by the potential privacy risks due to the collection of sensitive user data (e.g., location, in-person contact events). In this project, we analyze the privacy and utility trade-offs drawing on existing app design proposals, and conduct a survey study to understand how general people think about them and what are their preferred solutions. Our results may shed light on how to design a COVID-19 contact tracing app that can strike a good balance between the benefits of public health and the risks of security and privacy of user data, and provide insights about how to communicate with the general public about the risks and benefits in this special type of app.

We are currently at the stage of distributing the surveys and analyzing the results. We will update our findings on this website once they are ready to publish.

Contact tracing is a key strategy for preventing futher spread of COVID-19. In contact tracing, public health staff work with a patient to help them recall everyone with whom they have had close contact during the timeframe while they may have been infectious.

Contact tracing app is a smartphone technology that are built to facilitate health authorities with the contact tracing process. It can track either where a person has been, or who they have been in close contact with (e.g., within 2 meters/6 feet for more than 15 minutes), or both, within a recent period of time. When someone is diagnosed as being infected, all close contacts in the past 14 days (i.e., the incubation period of the virus) tracked by this system can be alerted immediately, which could enable the person to receive tests and go into self-quarantine in a timely fashion. The system may also help monitor the early stages of local outbreaks and give precaution alerts to the general public.

People’s perceptions of the risks and benefits of a COVID-19 contact tracing app can directly affect their intention about installing it. As suggested by theoretical modeling and the experience of some early-adopters of digital contact tracing apps (e.g., Singapore), obtaining a critical mass of installation is one of the key requirements to the success of app-based contact tracing systems that rely on voluntary participation. Therefore, there is a pressing need to understand how the general public think about contact tracing apps so that effective contact tracing apps can be built and deployed to curb the spread of the disease.

There is an intrinsic tension between the effectiveness of these apps and the potential invasion of users’ privacy and civil liberties. For example, showing detailed location information of people who are infected with COVID-19 can help people figure out whether they may have been exposed to the virus, while this transparency also increases the risks of disclosing the identities of COVID-19 patients to the public. We already observed multiple directions in handling the trade-offs among different aspects of requirements in existing systems. However, there has yet to be a "perfect solution" -- namely, that the gain of one aspect often comes at the cost of the loss of another aspect.

By articulating the risks and benefits in this type of apps proactively and honestly to a general audience, we can gather crucial feedback from them before actually building and deploying the app. Our results would serve as a solid foundation to help related stakeholders make the right decisions for their communities (e.g., federal/state-level health authorities that lead the development of the app, big tech companies that provide technical support for developing the mobile contact tracing apps). This can also help us gain an estimation of how people react to these apps if they are well-informed of their pros and cons.

We are currently conducting a survey study on Amazon Mechanical Turk to elicit opinions about different designs of contact tracing apps, targeting a general audience (currently limited to participants in the United States). To inform the design of the survey, we have run a comprehensive analysis of existing contact tracing systems to identify common patterns in their designs and trade-offs. This analysis covers both apps that have been deployed in certain countries, and proposals and prototypes from research institutes and companies (such as Google and Apple).

Our team consists of researchers from Carnegie Mellon University and Stanford University specializing in multiple areas, including computer science, human-computer interaction, privacy, social computing, and public policy.

Team members: Tianshi Li (Carnegie Mellon University), Jackie (Junrui) Yang (Stanford University), Cori Faklaris (Carnegie Mellon University), Jen King (Stanford University), Yuvraj Agarwal (Carnegie Mellon University), Laura Dabbish (Carnegie Mellon University), Jason I. Hong (Carnegie Mellon University)

We are looking forward to collaborating with researchers in various fields around the world (e.g., computer security and privacy, human-computer interaction, design, public policy, public health, epidemiology, etc.).