Fake “like” factories – how we reverse engineered facebooks user IDs [video]

Re:Fake “like” factories – how we reverse engineered faceboo... by stevenicr on Monday December 30, 2019 @11:51PM (#21919049)

This is terrible. Someone with access to server logs at a few web sites (or someone that loads a third party resource (ads, analytics, sharethis buttons, etc) onto a few web sites) would be able to know exactly who is coming from fbook and viewing different types of content.. no inside access to fbook or their data required.

I can't believe nothing is auto-stripping this url addition between firefox and ublock origin - must find and tell others.

[ Parent | on HN ]

Re:Fake “like” factories – how we reverse engineered faceboo... by _-___________-_ on Tuesday December 31, 2019 @05:56AM (#21920190)

Note that the ID changes for each "click" - hence "Facebook CLICK ID" - so you can't tie different clicks back to the same user without access to Facebook's data.

But yes, an extension to strip these (along with `gclid` from Google, `reddit_cid` from Reddit, and so on) would be very welcome.

[ Parent | on HN ]

Re:Fake “like” factories – how we reverse engineered faceboo... by stevenicr on Tuesday December 31, 2019 @06:52AM (#21920438)

thanks for this - I was confused about it since I have seen a few that looked very similar (first 12 characters of it the same) - did a quick, slightly deeper check into history and see that I do have 2 urls that have the exact same string, but it appears at least in this one case, that it's a click to a short link that forwards (dj-m.ag to djmag dot com) - so the similarity was confusing, and see some doubles freaked me out. now I see at least some differences in several - so I guess unless there is an easy pattern match, or as the title of this hn thread story suggests they may be able to reverse engineer fbook clicks - it's highly concerning to be sure since so many places use third party scripts and assets.

fingers crossed this is good and it gets better.

[ Parent | on HN ]