Facing Real-Time Identification in Mobile Apps & Wearable Computers

Article excerpt

TABLE OF CONTENTS  INTRODUCTION I. REAL-TIME FACE RECOGNITION TECHNOLOGY USING    PHONES (AND GLASSES)    A. The Process of Automatic Face Recognition in Real       Time    B. Early Applications of Real-Time Identification    C. Cyborgs, Wearable Computers, and Augmented       Reality II. CONCEPTUAL SIMILARITIES (AND DIFFERENCES) BETWEEN     REAL-TIME IDENTIFICATION AND GEOLOCATION     APPLICATIONS    A. Location, Location, Location    B. No Notice or Consent    C. The Ability to De-Anonymize a Face    D. Government Access to Data III. EMERGING REGULATORY RESPONSES TO FACE      RECOGNITION TECHNOLOGY    A. Federal Trade Commission Guidelines on Face       Recognition Technology    B. European Union Article 29 Working Party Opinion       on Facial Recognition in Online and Mobile Services IV. INITIAL POLICY RECOMMENDATIONS FOR REAL-TIME     IDENTIFICATION    A. Focus on Use Rather than Technology    B. Security by Design    C. Ask (the Right Person) for Permission    D. Collect Less; Delete More    E. Think About the Context and User Experience       Design CONCLUSION 

INTRODUCTION

In a thrilling scene in the computer-animated film The Incredibles, Mr. Incredible stumbles upon a tablet-like device. The device scans his face with a camera, identifies him as Mr. Incredible, and proceeds with telling him a classified message before it self-destructs. Is this technology something you would only see in a fiction cartoon about superheroes? As it happens, it is neither imaginary nor sci-fi. In fact, a mobile application with similar functionality can today be downloaded instantly to your smartphone for $1.99--save the self-destruction. (1)

But while we may observe some mobile applications of face recognition technology crop up in the iTunes store and elsewhere, this technology is still in its infancy. Computer scientists have been working on face recognition technology for decades, but the technology has only recently been implemented in consumer applications. These applications leverage the vast amount of labeled photos aggregated in social networks and the users' oblivious keenness to teach algorithms how to recognize their friends. The ubiquity of mobile phones with built-in cameras presents a new opportunity for this technology. For now, face recognition with mobile phones requires fast Internet connection to communicate with servers that can store all the data about faces and process the information. (2) But this too is now being enabled through rapid progress in mobile Internet speeds and the deployment of 4G mobile broadband. (3)

The use of face recognition technology in mobile apps challenges individuals' ability to remain anonymous in public places. These apps--in their current iteration--encourage users to upload photos with identified faces to social networks, along with embedded metadata revealing where and when they were captured. Consequently, when uploaded, the labeled images generate a digital paper trail of the individuals' location in the photo. The apps have this effect without seeking the consent of the identified individuals, who will not have seen the privacy notice displayed when an app was downloaded, and may not even know that they were photographed or identified. In essence, this practice subjects individuals to a possible surveillance by their peers, employers, and companies that have an interest in their everyday choices, and perhaps even the government. The ability to go off the radar allows for quiet reflection and daring experimentation--processes that are essential to a productive and democratic society. (4) In the words of privacy scholar Julie Cohen, "citizens who are subject to pervasively distributed surveillance and modulation by powerful commercial and political interests... increasingly will lack the capacity to form and pursue meaningful agendas for human flourishing." (5) Given what we stand to lose, we ought to be cautious with groundbreaking technological progress. …