Browser plugin: fuzzify.me
Illustrations by Clara Juliano for Coding Rights
fuzzify.me (2018) is a (now defunct) browser extension available for both Chrome and Firefox that helps give Facebook users more insight into how advertisers on the platform might be targeting them. The tool allows users to automatically and routinely deny advertisers access to their personal information, while also showing them a complete timeline of how they are being targeted over time.
The project was a collaboration between myself, Hang Do Thi Duc, and Joana Varon of Coding Rights as part of our time as 2017-18 Mozilla Fellows. We worked together on both the development of the tool, which can be accessed here, and accompanying research into how advertisements on Facebook often target women in discriminatory ways.
After installation, the browser extension begins assembling a running list of all the Facebook advertisements the Facebook user is seeing over time. The user can view all those advertisements in the Fuzzify.me dashboard, as well as the explanation Facebook provides for why they are seeing that advertisement. All the information is saved locally, so only the user has access to the information the browser extension collects.
Besides providing greater transparency, the tool also helps Facebook users minimize the information advertisers access through the Facebook ad platform. Facebook regularly assigns users “ads categories,” which can be viewed in the ad preferences page on Facebook. From the Fuzzify.me dashboard, the user can click “Clean Ads,” automatically removing all the categories that are visible to advertisers. The categories you cleared from Facebook is also logged into the user dashboard.
The idea is that over time users will able to compare how their targeted ads are changing over time after making changes to their advertising preferences. Ultimately, the goal is to investigate whether or not the privacy controls Facebook has implemented are actually effective at preventing or minimizing ad targeting.
For more information, you can see the code on Github.