Formal Privacy Analyses for Open Banking

The term “Open Banking” describes a series of global initiatives to allow the sharing of customer data between financial companies to facilitate competition within their sector. In this paper, we formalise in the rigorous framework of quantitative information flow (QIF) relevant privacy risks in a concrete Open Banking scenario, namely: (i) transaction-history recovery and (ii) collateral attribute-inferences using external correlations. We provide extensive analyses of these risks in real-world data from Open Banking, supplied by a fintech in Australia. We show that the Open Banking system studied presents considerable privacy risks with respect to transactions, both in the presence and in the absence of demographic data. Finally, we exemplify potential real-world collateral attribute-inference attacks, in which we show how an attacker might leverage scientific correlations to infer individuals’ level of neuroticism and self-control from their transaction history. We hope that this work may: (i) help financial customers in Australia make better-informed decisions about what kind of information, and how much of it, to share via Open Banking; (ii) raise awareness about the potential privacy risks of Open Banking in other countries; and (iii) foster the development of privacy regulation in digital finance and the open data economy.