Pool Inference Attacks on Local Differential Privacy: Quantifying the Privacy Guarantees of Apple's Count Mean Sketch in Practice

TL;DR

This paper introduces pool inference attacks against Apple's Count Mean Sketch under local differential privacy, demonstrating how polarized user behaviors can be exploited to infer sensitive preferences, especially with high privacy budgets.

Contribution

The paper presents the first formalization of pool inference attacks on local differential privacy mechanisms like Apple's Count Mean Sketch, with practical evaluations on emoji and web browsing data.

Findings

Attacks outperform random guessing with sufficient data collection.

Users with high polarization are more vulnerable to inference.

The attack is validated using real Twitter user data.

Abstract

Behavioral data generated by users' devices, ranging from emoji use to pages visited, are collected at scale to improve apps and services. These data, however, contain fine-grained records and can reveal sensitive information about individual users. Local differential privacy has been used by companies as a solution to collect data from users while preserving privacy. We here first introduce pool inference attacks, where an adversary has access to a user's obfuscated data, defines pools of objects, and exploits the user's polarized behavior in multiple data collections to infer the user's preferred pool. Second, we instantiate this attack against Count Mean Sketch, a local differential privacy mechanism proposed by Apple and deployed in iOS and Mac OS devices, using a Bayesian model. Using Apple's parameters for the privacy loss $\varepsilon$, we then consider two specific attacks: one in the emojis setting -- where an adversary aims at inferring a user's preferred skin tone for emojis -- and one against visited websites -- where an adversary wants to learn the political orientation of a user from the news websites they visit. In both cases, we show the attack to be much more effective than a random guess when the adversary collects enough data. We find that users with high polarization and relevant interest are significantly more vulnerable, and we show that our attack is well-calibrated, allowing the adversary to target such vulnerable users. We finally validate our results for the emojis setting using user data from Twitter. Taken together, our results show that pool inference attacks are a concern for data protected by local differential privacy mechanisms with a large $\varepsilon$, emphasizing the need for additional technical safeguards and the need for more research on how to apply local differential privacy for multiple collections.