[Paper Review] Privacy Guidelines for Contact Tracing Applications

Manish Shukla, Rajan M. A+3 more|arXiv (Cornell University)|Apr 28, 2020

COVID-19 Digital Contact Tracing15 references21 citations

TL;DR

This paper proposes comprehensive privacy guidelines for contact tracing applications to balance public health needs with user data protection. It outlines stakeholder-specific recommendations for developers, governments, and businesses, emphasizing decentralized data processing, end-to-end encryption, transparency, and usability to ensure trust and mass adoption during infectious disease outbreaks like COVID-19.

ABSTRACT

Contact tracing is a very powerful method to implement and enforce social distancing to avoid spreading of infectious diseases. The traditional approach of contact tracing is time consuming, manpower intensive, dangerous and prone to error due to fatigue or lack of skill. Due to this there is an emergence of mobile based applications for contact tracing. These applications primarily utilize a combination of GPS based absolute location and Bluetooth based relative location remitted from user's smartphone to infer various insights. These applications have eased the task of contact tracing; however, they also have severe implication on user's privacy, for example, mass surveillance, personal information leakage and additionally revealing the behavioral patterns of the user. This impact on user's privacy leads to trust deficit in these applications, and hence defeats their purpose. In this work we discuss the various scenarios which a contact tracing application should be able to handle. We highlight the privacy handling of some of the prominent contact tracing applications. Additionally, we describe the various threat actors who can disrupt its working, or misuse end user's data, or hamper its mass adoption. Finally, we present privacy guidelines for contact tracing applications from different stakeholder's perspective. To best of our knowledge, this is the first generic work which provides privacy guidelines for contact tracing applications.

Motivation & Objective

• Address growing privacy concerns in mobile contact tracing applications during infectious disease outbreaks.
• Identify threats to user privacy, including surveillance, data leakage, and behavioral pattern exposure.
• Provide actionable, stakeholder-specific privacy guidelines for developers, governments, and businesses.
• Ensure mass adoption by balancing public health utility with strong data protection and user trust.
• Establish a framework for secure, transparent, and privacy-preserving contact tracing systems.

Proposed method

• Analyze privacy controls in prominent contact tracing applications (e.g., Singapore’s TraceTogether, Israel’s app) to identify strengths and weaknesses.
• Propose a multi-stakeholder privacy framework covering developers, governments, and businesses.
• Advocate for decentralized data architecture to minimize central data storage and reduce breach risks.
• Recommend use of privacy-preserving techniques such as private set intersection and bloom filters for secure computation.
• Emphasize end-to-end encryption, data anonymization, and secure local storage of sensitive data on user devices.
• Call for open-sourcing the application code and verifiable compilation to ensure transparency and prevent backdoors.

Experimental results

Research questions

• RQ1 How can contact tracing applications protect user privacy while remaining effective in curbing disease spread?
• RQ2 What are the key privacy threats posed by centralized data collection in contact tracing systems?
• RQ3 How can transparency and explainability in risk scoring algorithms increase user trust and adoption?
• RQ4 What technical and policy measures can prevent misuse of contact tracing data by third parties or state actors?
• RQ5 How can usability be optimized for diverse populations, including elderly and low-literacy users, without compromising privacy?

Key findings

• The paper identifies that centralized data collection in contact tracing apps poses significant privacy risks, including mass surveillance and data leakage.
• Decentralized architectures, where data is stored locally on devices and only anonymized risk scores are shared, significantly reduce privacy exposure.
• Open-sourcing the application code and verifying compiled binaries against source code enhances trust and enables third-party auditability.
• Providing clear, explainable risk scores with factor-level breakdowns increases user understanding and acceptance of the system.
• The inclusion of features like location blacklisting (e.g., home, office) allows users to control data exposure and improves usability.
• Proper data retention policies—such as automatic deletion of infected contact trails after quarantine periods—help prevent long-term privacy violations.