Beyond Anonymisation – Anonymising Data

Anonymisation is often considered the panacea of data protection: it helps organisations derive useful insights from data while allowing them to avoid the legal and privacy risks associated with processing personal data… or so the theory goes.

However, as a standalone measure anonymisation can fall short. To cite a famous example, a graduate student was once able to use nominally anonymised hospital data from Massachusetts to send the state’s Governor his own health records. We should therefore be careful about the use of the word ‘anonymised’. As the UK’s Information Commissioner’s Office (ICO) points out, it risks leaving organisations with the false sense of security that data is impossible to re-identify under any circumstances.

In this whitepaper we want to build on the UK’s Understanding Patient Data’s reconceptualisation of anonymisation as ‘de-personalisation’. First, we explore a range of techniques that enable organisations to minimise the exposure of personal data. We’ll argue that anonymisation redefined as such can reduce the risk associated with data processing but does not necessarily absolve organisations of their data protection responsibilities – nor need that be the primary goal. We’ll then demonstrate how our customers can safely process personal data in Palantir Foundry by relying on a range of built-in privacy-protective means.