Article by Chen Sun, Nikolas Guggenberger, and Supreeth Shastri: “Ancient wisdom says that everything that has a beginning has an ending. However, when it comes to the lifecycle of personal data, the ending is nowhere in sight. In fact, for much of its existence, the computing community has evolved without treating deletion as a first-class operation. This practice had to change when the European Union introduced Right to be Forgotten (RTBF)—first in 2014, as a standalone right applicable to online search engines and then in 2018, as a universal right applicable to all data controllers through the General Data Protection Regulation (GDPR).
“Isn’t it just deletion?” has been the computing community’s standard reaction to the RTBF’s requirements. While the end goal of RTBF is indeed the deletion of data, casting RTBF as just deletion is akin to saying that eating is just for nutrition. It is not surprising that over the first five years of GDPR, an RTBF penalty is issued once every nine days—a clear sign that the computing and data management communities have continued to oversimplify, misunderstand, and poorly implement RTBF. Our work is an attempt to remedy this disconnect.
This article demonstrates how RTBF exposes computing systems to uncertainties and the challenges at all stages of design and operation, and how RTBF has invalidated principles and practices of data management with decades of precedent. To address these challenges, we propose a principled approach for introducing RTBF capability in computing systems. Our solution is rooted in two key insights:
- The need to bridge an intrinsic dichotomy existing between computing and law, that is, computing systems are created to be precise and static, but laws are written to be abstract and interpretable
- Modeling compliance as a via negativa problem, that is, instead of trying to build a perfectly compliant system, it is much easier to weed out known violations from the system…(More)”.
