The End of Privacy as We Know It

Samuel Brice
3 min readNov 7, 2020

--

Demystifying Clearview AI Blog Series (Introduction)

Table of Contents

Previous: Demystifying Clearview AI

Next: A Short History of Facial Recognition

Photo by Keiligh Baker

The New York Times was the first to publish an in-depth report about Clearview AI — the secretive company that may “end privacy as we know it.” As described in the article, the startup claims to help law enforcement match photos of unknown people to their online images — and “might lead to a dystopian future or something,” a backer said.

Shortly after the New York Times article, more news outlets started publishing details about the billions of photos Clearview collected from sites such as Facebook and Instagram, as well as the broad range of clients that have been actively using Clearview’s facial recognition app for various purposes.

A Buzzfeed review of Clearview’s leaked client list revealed that as many as 2,228 law enforcement agencies, companies, and institutions had collectively performed nearly 500,000 searches using the app. In addition to federal agencies such as the Justice Department and state agencies such as the New York State Police, Clearview’s client list also includes international entities and private individuals such as the United Arad Emirates and Clearview’s angel investors.

Whereas most articles have focused primarily on What Clearview has collected and Who has been using the app, I’ll be focusing How Clearview’s Deep Learning technology works by demonstrating how you’d piece together a similar system using off-the-shelf free and open-source tools.

We’ll first start with a short history lesson about facial recognition technology. Then we’ll dive into the Deep Learning model development process.

As a concrete demonstration of the underlying technologies, we’ll go through the process of implementing a vehicle recognition app called CCTView using publicly available CCTV camera feeds from the New York City Department of Transportation.

Finally, we’ll discuss the high-level privacy implications of such technologies and explore some general tools for defending against such invasive technologies.

--

--

Samuel Brice
Samuel Brice

Written by Samuel Brice

This page intentionally left blank.

No responses yet