Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

You Still See Me: How Data Protection Supports the Architecture of AI Surveillance

Rui-Jie Yew, Lucy Qin, Suresh Venkatasubramanian

TL;DR

The role that technologists could play in devising policies that combat surveillance AI technologies is highlighted by highlighting the role that technologists could play in devising policies that combat surveillance AI technologies.

Abstract

Data forms the backbone of artificial intelligence (AI). Privacy and data protection laws thus have strong bearing on AI systems. Shielded by the rhetoric of compliance with data protection and privacy regulations, privacy-preserving techniques have enabled the extraction of more and new forms of data. We illustrate how the application of privacy-preserving techniques in the development of AI systems--from private set intersection as part of dataset curation to homomorphic encryption and federated learning as part of model computation--can further support surveillance infrastructure under the guise of regulatory permissibility. Finally, we propose technology and policy strategies to evaluate privacy-preserving techniques in light of the protections they actually confer. We conclude by highlighting the role that technologists could play in devising policies that combat surveillance AI technologies.

You Still See Me: How Data Protection Supports the Architecture of AI Surveillance

TL;DR

The role that technologists could play in devising policies that combat surveillance AI technologies is highlighted by highlighting the role that technologists could play in devising policies that combat surveillance AI technologies.

Abstract

Data forms the backbone of artificial intelligence (AI). Privacy and data protection laws thus have strong bearing on AI systems. Shielded by the rhetoric of compliance with data protection and privacy regulations, privacy-preserving techniques have enabled the extraction of more and new forms of data. We illustrate how the application of privacy-preserving techniques in the development of AI systems--from private set intersection as part of dataset curation to homomorphic encryption and federated learning as part of model computation--can further support surveillance infrastructure under the guise of regulatory permissibility. Finally, we propose technology and policy strategies to evaluate privacy-preserving techniques in light of the protections they actually confer. We conclude by highlighting the role that technologists could play in devising policies that combat surveillance AI technologies.
Paper Structure (24 sections, 1 figure)

This paper contains 24 sections, 1 figure.

Table of Contents

  1. Abstract
  2. Introduction
  3. Definitions and Background
  4. Privacy-Preserving AI
  5. Dataset Curation.
  6. Model Training and Computation.
  7. Using PPAI: Model Application.
  8. Policy Analysis: PPAI as Compliance with Data Protection and Privacy Laws
  9. Surveillance
  10. Privacy and Contextual Integrity
  11. Seeing Through the Data-Protected Layers of the PPAI Pipeline
  12. Producing PPAI: Dataset Curation (You see even more of me.)
  13. Producing PPAI: Model Training and Computation (You're protecting you.)
  14. Using PPAI: Model Application (Don't use them against me!)
  15. Regulating PPAI: Technology and Policy Strategies
  16. ...and 9 more sections

Figures (1)

  • Figure 1: This illustration is from liu2023vertical. It shows that Party A and Party B have the same samples, but different features of those samples. PSI, which is a pre-processing phase for vertical FL aligns the datasets from Party A and Party B to capture richer features for the samples that they share. Thus, for instance, if Party A has information that a data subject viewed an advertisement and Party B has information that a data subject bought the product served in the advertisement, then that information can be combined for training a resulting model.