A Framework for Measuring Privacy Risks of YouTube

Abstract

While privacy risks associated with known social networks such as Facebook and Instagram are well studied, there is a limited investigation of privacy risks of YouTube videos, which are mainly uploaded by teenagers and young adults, called YouTubers. This research aims on quantifying the privacy risks of videos when sensitive information about the private life of a YouTuber is being shared publicly. We developed a privacy metric for YouTube videos called Privacy Exposure Index (PEI) extending the existing social networking privacy frameworks. To understand the factors moderating privacy behaviour of YouTubers, we conducted an extensive survey of about 100 YouTubers. We have also investigated how YouTube Subscribers and Viewers may desire to influence the privacy exposure of YouTubers through interactive commenting on Videos or using other parallels YouTubers’ social networking channels. For this purpose, we conducted a second survey of about 2000 viewers. The results of these surveys demonstrate that YouTubers are concerned about their privacy. Nevertheless inconsistent to this concern they exhibit privacy exposing behaviour on their videos. In addition, we found YouTubers are being encouraged by their audience to continue disclosing more personal information on new contents. Finally, we empirically evaluated the soundness, consistency and applicability of PEI by analyzing 100 videos uploaded by 10 YouTubers over a period of two years.