How FB uses fingerprint to detect you06.10.2017Reading Time: 3 minutes
Until recently, there was an opinion that Facebook does not collect any data about the platform and device with which the affiliate is working, because the browser itself does not transmit such data to the network, and Facebook can collect this information only with the help of third-party services and applications.
Let me make it crystal clear: Facebook collects, accesses, processes, stores and segment the information about your device. Perhaps part of the data received by the social network hadn’t been used until recently, but the fact is: FB has the ability to extract the information about the hardware installed on the device and some of the system settings.
Since April, FB launched the trust-rate algorithms for accounts on the principle of Fingerprint (technologies that Google has been actively using for more than a year); According to the new algorithm all new accounts created using cringed or suspicious resources uncompromisingly are sent to the so-called “Selfie” checkpoint. On one hand, it’s good that they are not immediately blocked. On the other hand, getting through this algorithm does not guarantee that you will not end up there again hours later. Machine learning from Facebook is ruthless: the algorithm will reset the same account over and over again if the fingerprint of the profile seems suspicious.
Let’s get to the point. There is an alternative user tracking method beside cookies. If you turn off cookies in the browser (by the way FB inserts 11 different cookies in your browser) your profile immediately ends up in the suspicious list. This significantly lowers the trust-rate, which makes it harder to protect yourself from fingerprint technology. The way it works: js-scripts consistently poll your browser for a whole heap of very platform-dependent settings. We call this technology Fingerprint-js, but Facebook uses libraries of its own development. If you want, you can debug all FB scripts and find all such requests.
Using fingerprint technology, FB recognizes not only the “stuffing” of your computer, but also indirectly learns information about installed libraries, software, drivers, etc. Each electronic device that got in facebook’s line of sight is assigned a unique digital label in the form of a hash sum. Add this to the data collected from double entry technology, data collected with cookies, and data that FB buys from third-party services. Daily FB receives roughly 1000 terabytes of information about its users, which social network engineers systematize, structure and analyze with the help of machine learning.
Not long ago we diligently tried to hide from the Network. Webrtc and flash partially blocked outgoing traffic, blocked cookies, cleaned cache and at the same time made us feel smart. Today, new registration with such parameters will directly go to the Selfie page faster than you can pronounce the name of FB owner. Previously, we were satisfied when the bunny whoer showed an honest 100% of anonymity. Now times have changed, and FB freaks out just like Google in 2015.
Today we have to be even smarter. Blocking cookies is just as good as confessing to the network – hey, look, I’m an arbitrator, I’m here. The truth is that 95% of network users do not know what cookies are and what they do. At the same time everyone who is trying to hide looks suspicious. And for FB suspicious – means guilty. This leads to either your account being “flagged” or to the good old Selfie check. You have to work really hard to convince FB that you are an average user. Bans in FB always were and remain a scoring system, but still nobody knows the variables in the trust rate calculation formula.
Back to the point: How much does Facebook know about the webmaster? A LOT! In reality it is much more than we think.