Table of contents

Omer Tene, Katrine Evans, Bruno Gencarelli, Gabe Maldoff, Gabriela Zanfir-Fortuna, Nurul Momen, Majid Hatamian, Lothar Fritsch, Jatinder Singh, Jennifer Cobbe, Cesare Bartolini, Gabriele Lenzini, Livio Robaldo, Bülent Yener, Tsvi Gal, M. Sadegh Riazi, Darvish Bita, Farinaz Rouhani, Dan Boneh, Andrew Grotto, Patrick McDaniel, Nicolas Papernot, Scott Ruoti, Kent Seamons, Alan T. Sherman, Linda Oliva, Enis Golaszewski, Dhananjay S. Phatak, Travis Scheponik, Geoffrey Herman, Dong Wik Choi, Spencer Offenberger, Peter A. Petérson, Josiah Dykstra, Gregory V. Bard, Ankur Chattopadhyay, Filipo Sharevski, R. C. Verma, Ryan Vrecenar, Andrew Fasano, Tim Leek, Brendan Dolan-Gavitt, Joshua Bundt, James Michael, Daniel E. Geer, Dale PetersonIEEE Security & Privacy 2019 · DOI

Abstract

Nearly a decade in the making, the General Data Protection Regulation (GDPR), Europe's massive overhaul of its privacy and data protection laws, came into effect to great fanfare in May 2018. Impacting every area of an economy marked by technological and data innovation, including the public and private sectors, finance and health care, retail and education, transportation, pharmaceuticals, utilities, and scientific research, the GDPR carried immense promise but also many implementation challenges and interpretation complexities. A year in, it is not too soon to pause for reflection and to explore the reform's effect on corporate and organizational data practices, particularly at the intersection of policy, law, and engineering.

Introduction: The Translation Gap

While the GDPR was a landmark legal overhaul, its implementation created a massive "translation" gap between high-level policy and low-level technical execution. Engineers and data scientists were suddenly tasked with building systems that guarantee "privacy by design," the "right to be forgotten," and "data portability" without a clear technical roadmap. This document, based on the November/December 2019 issue of IEEE Security & Privacy, evaluates the impact of the regulation shortly after its first anniversary, moving beyond legal theory to examine how the regulation has fundamentally reshaped the intersection of policy, law, and engineering.

The GDPR requires a fundamental shift in how data layers are accessed and protected.

The Engineering Phase of Privacy

The guest editors of this special issue argue that we have moved past the "legal drafting" phase of GDPR and into the "engineering" phase. In this new era, the success of the law depends on those who build the systems. Compliance is no longer a "check-the-box" exercise but a formal model that must be integrated into the software development lifecycle. Data Protection Authorities (DPAs) are increasingly looking at the technical implementation of privacy, not just the legal privacy policy. To assist in this, researchers have introduced the DAta Protection REgulation COmpliance Model (DA-RE), which suggests a structured, logic-based approach to mapping legal requirements to organizational processes.

Security Risks of Data Rights

A counter-intuitive insight from the first year of implementation is that granting users more rights—such as the right to access all their data (Data Subject Access Rights)—creates new security risks. If an authentication process for a data request is weak, the GDPR itself could become a tool for identity theft or social engineering. Security threat modeling is now essential to ensure that these new rights do not become weaponized attack vectors.

Empirical Realities and Technical Solutions

Empirical analysis of the mobile app ecosystem suggests that while the GDPR exists, the actual improvement in app privacy practices is a complex, ongoing process rather than an overnight success. Research by Momen et al. indicates that data collection practices have not tangibly changed in all sectors, highlighting the difficulty of measuring whether privacy-focused laws result in better privacy for end-users.

However, technical solutions are emerging to meet these challenges. The issue highlights advancements in Privacy-Preserving Machine Learning (PPML), suggesting that techniques like homomorphic encryption and differential privacy are becoming essential for GDPR-compliant AI. These methods allow for deep learning on private data without compromising individual identities.

Future Challenges: AI, Bots, and Usability

As we look beyond the first year, new challenges arise at the intersection of AI and regulation. The rise of "Sophisbots"—AI agents so sophisticated they may render traditional Turing Tests irrelevant—necessitates new ways to verify human versus bot interactions in a regulated environment.

The intersection of AI, automated bots, and human data processing presents new regulatory hurdles.

Furthermore, the "sociotechnical" aspect of security remains a hurdle. Technical compliance is ineffective if the resulting systems, such as secure email, are not usable for the average person. The long-term effects of GDPR on innovation and the final "settled" interpretations of the law by courts will continue to evolve, requiring ongoing collaboration between legal experts and system engineers.

Summary updated

Abstract

Nearly a decade in the making, the General Data Protection Regulation (GDPR), Europe's massive overhaul of its privacy and data protection laws, came into effect to great fanfare in May 2018. Impacting every area of an economy marked by technological and data innovation, including the public and private sectors, finance and health care, retail and education, transportation, pharmaceuticals, utilities, and scientific research, the GDPR carried immense promise but also many implementation challenges and interpretation complexities. A year in, it is not too soon to pause for reflection and to explore the reform's effect on corporate and organizational data practices, particularly at the intersection of policy, law, and engineering.
BibTeX
@inproceedings{tene2019table,
  author = {Omer Tene and Katrine Evans and Bruno Gencarelli and Gabe Maldoff and Gabriela Zanfir-Fortuna and Nurul Momen and Majid Hatamian and Lothar Fritsch and Jatinder Singh and Jennifer Cobbe and Cesare Bartolini and Gabriele Lenzini and Livio Robaldo and Bülent Yener and Tsvi Gal and M. Sadegh Riazi and Darvish Bita and Farinaz Rouhani and Dan Boneh and Andrew Grotto and Patrick McDaniel and Nicolas Papernot and Scott Ruoti and Kent Seamons and Alan T. Sherman and Linda Oliva and Enis Golaszewski and Dhananjay S. Phatak and Travis Scheponik and Geoffrey Herman and Dong Wik Choi and Spencer Offenberger and Peter A. Petérson and Josiah Dykstra and Gregory V. Bard and Ankur Chattopadhyay and Filipo Sharevski and R. C. Verma and Ryan Vrecenar and Andrew Fasano and Tim Leek and Brendan Dolan-Gavitt and Joshua Bundt and James Michael and Daniel E. Geer and Dale Peterson and Daniel Geer and Dale Peterson},
  title = {{Table of contents}},
  booktitle = {IEEE Security \& Privacy},
  year = {2019},
  doi = {10.1109/msec.2019.2936696},
}
Figure