Mozilla Fights Online Tracking with "Collusion"

The Mozilla Foundation announced yesterday the release of a new Firefox add-on intent on helping users understand how tracking cookies can be used by advertisers and others to assemble a nearly comprehensive picture of users' activity online. Collusion is the latest tool released in Mozilla's ongoing effort to empower users to have more control over how they interact with the web. Its current implementation is focused on a graph representation of the flow of cookies, where nodes represent unique domains and edges represent that a request was made from a property of the first domain to one of the second. A demo of this visualization in action can be found on the above-linked Collusion page.

The idea is that within just a handful of visits to supposedly independent websites, it becomes quite clear that several advertisers have a very good idea of exactly how a given user interacts with the web. This is, of course, a manifestation of a couple of the universal properties of graph structure that we have discussed in class, in that the graph quickly converges to a giant strongly connected component with high connectivity. Mozilla goes one step further in the actual add-on by surrounding nodes that a user has explicitly requested with a white halo, and those that are "confirmed trackers by privacychoice.org" with a red one. Privacy advocates at Mozilla hope that the striking visualization will serve as a concerning wakeup call to web users heretofore unaware of the nature of tracking cookies. Empowering users to do something about it is mostly left for a later version. They state that future versions of the Collusion add-on will include the ability to block individual trackers as well as an opt-in to anonymously provide data to researchers and privacy advocates interested in investigating aggregate trends in internet tracking.

If this initiative, along with other Mozilla privacy initiatives like Do Not Track, is successful in encouraging a non-trivial portion of web users to take active control of their privacy on the web, it will be interesting to see what effects it has on the internet advertising industry. As we saw in our in-class experiment, this kind of rich data about user activity is extremely valuable to advertisers, and users opting out may make internet advertising a less lucrative game.