Natasha Fernandes, Gabriel H. Nunes, Annabelle McIver
Quantitative Information Flow (QIF) is a decision- and information-theoretic framework based on Formal Methods that facilitates the analysis of complex computational systems. The Topics API is being proposed by Google as a more privacy-friendly alternative to third-party cookies for Interest-Based Advertisement (IBA). Third-party cookies allow the precise tracking of individuals’ Internet browsing histories. The Topics API represents an individual as a set of top topics of interest derived from their browsing history and a topics taxonomy. API callers can learn those topics or a random topic from the whole taxonomy with 5% chance. We use QIF to precisely measure the privacy and utility effects of each aspect of the Topics API, and to verify Google’s privacy-related claims.