Skip to content

Personal data obfuscation for privacy

Ben Mezger edited this page Dec 5, 2016 · 2 revisions

📝 Editor's note: There's some belief that obfuscating personal data, such as search queries or clicking on ads, helps protect a user's privacy by hiding real actions inside of "ghost" (faked, automated) actions. The effectiveness of these techniques, despite being widely-touted, remain controversial.

This page can be seen as a supplement to the Persona-based commsec training matrix for those who want to employ this technique, but remains separated from that guide due to the specialized nature of this tactic and the lack of consensus or specificity on whether it is, in fact, privacy-enhancing under what circumstances.

Online services such as Facebook/Google/Apple/etc are constantly watching everything you do, so they can sell your data and targeted ads to third parties/governments. There are a few ways you can generate "noise" and confuse machine algorithms into making a wrong targeted profiling.

Note: Depending on (and from) what you are obfuscating, it requires time. For example, you cannot obfuscates data analysis from Facebook in one hour. It may require several days, months, and maybe years.

  • If you use Facebook or any other social media, make sure you post random garbage that has nothing to do with you (your personality, or anything that can profile you);
  • Use different (random) Firefox profiles.
  • Run a tor relay while using Tor or navigating through the web. If you do so, the messages coming out of your computer can be yours or may be just one among many that you are passing on for other people. Warning: this can also decrease your anonymity. It's a 50/50 chance. See this article for more information.

Helpful tools

  • Use AdNauseam to click all ads found in a page, so user profiling, targeting and surveillance becomes futile.
  • Use TrackMeNot to generate random search queries (in background). It hides users' actual search trails in a cloud of 'ghost' queries, significantly increasing the difficulty of aggregating such data into accurate or identifying user profiles.
Clone this wiki locally