Socialbots: Infiltrating Your Social Networking Sites

By creating 120 ‘fake’ Facebook users, referred to as Socialbots Ph.D. student Yazan Boshmaf and Dr. Beznosov, along with other researchers from ECE’s  NetSys Lab and LERSSE were able to harvest 46,500 email addresses and 14,500 physical addresses from users’ profiles. The Socialbots began contacting Facebook users by making random friend requests and were accepted by 1 in 5 people they contacted. Once the Socialbot made some “friends” it increased its number of friends by using the social networks of those users who accepted it, befriending the friends of friends. These friend requests were far more likely to be accepted; the research team found that 60% of these contacts were successful. Facebook’s “Facebook Immune System”, the massive security system protecting users, did not detect the infiltration of Socialbots.

The image above illustrates the fascinating findings of the Socialbot research. Each red dot, representing one of the Socialbots, has been accepted by a varying number of users (blue dots), selected at random to initially be approached by the Socialbots. The green dots are friends of the blue dots who were successfully befriended by the Socialbots. The lines between the dots represent friendships. As illustrated by this depiction, the “fake” Facebook users have been adept at creating online friendships through mutual connections.

This research has important repercussions, especially given the sizeable amount of information individuals divulge on social networking sites. Socialbots are successful at large-scale infiltration of social networking sites, as they can effectively imitate real social network users. Given the large amount of information given and time spent on these sites, social networking websites can be a virtual goldmine of material for those able to access it. Evidently, Dr. Beznosov and Boshmaf’s research has real-world implications.

The Socialbot Nets research will be presented by Yazan Boshmaf  at the Annual Computer Security Applications Conference (ACSAC) in December.

Read more:

The Socialbot Network: When Bots Socialize for Fame and Money by Yazan Boshmaf, Ildar Muslukhov, Konstantin Beznosov, Matei Ripeanu

New ‘Socialbot’ threat steals private info from Facebook – are YOUR friends what they seem? from the Daily Mail

Facebook easily infiltrated, mined for personal info from CBC.ca

New Scientist magazine