Wednesday, December 7, 2011

Facebook Keeps Shadow Profiles on Non-Users

This article started out as a report on the supposed shadow profiles that Facebook keeps on non-users. The arc of the story is clear: Some people choose not to have Facebook accounts because they don't trust the company or its web site, but if Facebook keeps secret data caches on non-users then it is up to something more than just providing a service. Here are some of the stories:

Facebook Is Building Shadow Profiles of Non-Users

and the actual document of the complaint:

Facebook Building 'Shadow Profiles' of Non-Members, Experts Allege

Facebook is 'building shadow profiles of NON-users', says complaint to privacy watchdog

[Whoops, that was two Murdoch links back to back. Here is an independent voice.]

Facebook Ireland accused of creating 'shadow profiles' on users, nonusers

Of course, Facebook says that it is not keeping profiles of non-users. It does admit to keeping names and emails and linking them to people who are users.

Facebook denies “Shadow Profile” claims; Risks €100k privacy fine

I started thinking about Facebook's claim. If they only keep name and email, that is more than most us would like, but it could be a lot less than all the rest of these background investigation sites have:

pipl.com
www.intelius.com
www.spokeo.com
www.peoplefinders.com
www.anywho.com
www.whowhere.com
www.zabasearch.com
www.peoplesmart.com
www.beenverified.com
backgroundsearch.com
www.peoplerecords.com
www.backgroundpi.com
www.snoopstation.com

Every single one of these sites keeps a "shadow profile" on millions of people. So is Facebook any worse? Or should we be just as alarmed at spokeo or pipl or snoopstation?

One thing that is different is that Facebook keeps social network data. Spokeo may not have any links data connecting Smith with Johnson. Or they might. But we know for sure that Facebook does, and that its plans for using network data are more sophisticated than what we've seen from the public records publishers in the past.

No comments:

Post a Comment