Section: New Results

Selling Off Privacy at Auction

Participants : Claude Castelluccia, Lukasz Olejnik, Minh-Dung Tran.

The first one is a privacy analysis of Real-Time Bidding (RTB) and Cookie Matching (CM). RTB is a technology that allows ad buyers (advertisers) and ad sellers (publishers) to buy and sell ad spaces at real-time auctions through ad exchanges. In RTB, when user visits a publisher page, the ad impression (i.e. one ad display in an ad space) and the user information are immediately broadcast by the ad exchange to a number of bidders (i.e. advertisers or their representatives) for them to bid for the chance to serve ads to this user. CM protocol allows the ad exchange and the bidder to synchronize their cookies of the same user, thus facilitating their exchange of user data.

In [13] , we characterize and quantify the potential user web history leakage from ad exchanges to bidders in RTB as a result of exchanging user data. We also discuss and quantify the extent to which companies can potentially collude to increase their tracked user profiles using CM. In addition, we leverage a design characteristic of RTB to observe the winning price of each RTB auction. By analyzing these prices, we show how advertisers evaluate the value of user privacy. This work (titled Selling Off Privacy at Auction) will be presented in NDSS 2014, San Diego, USA in February, 2014.