11 Comments

Summary:

Facebook has been pilloried for not caring enough about our privacy. But now they face a call to offer data portability, something that could, if not carefully designed, compromise the privacy we so wanted last year. Facebook disabled blogger Robert Scoble’s account after he ran automated […]

Facebook has been pilloried for not caring enough about our privacy. But now they face a call to offer data portability, something that could, if not carefully designed, compromise the privacy we so wanted last year.

Facebook disabled blogger Robert Scoble’s account after he ran automated scripts against the site. The site’s Terms of Service say that you agree not to “use automated scripts to collect information from or otherwise interact with the Service or the Site.”

The general consensus seems to be that this was Scoble’s data and so he should be able to do whatever he likes with it. But that information he’s trying to get wasn’t all his. Apparently he wanted information about his “social graph”: the friendships he has recorded on Facebook and profile data about those friends.

Even if Scoble’s Facebook friends agreed to let him view their data on Facebook, they didn’t agree to let him take that information wherever he wants to do with what he wants. He could use a screen scraping program to grab data that they consider just-among-friends and stick it out in public without any regard for their privacy settings. You might say, “Scoble wouldn’t do that” but it’s Facebook’s responsibility to see that it doesn’t happen.

Data portability could be designed into Facebook in such a way that it doesn’t compromise user’s privacy. At the very least, an opt-in to profile sharing outside Facebook would need to be provided. Allowing uncontrolled screen scraping is not the answer.

You’re subscribed! If you like, you can update your settings

  1. Facebook has yet to understand that the privacy continuum ranges from deeply personal to widely public. I’d use Facebook more as a digital hub if it offered me a public presence or a way to export/synchronize Facebook data that I provide publicly. As it is, I feed my external sources into Facebook so that friends there can follow what I do on the wider web. If the inverse relationship were possible (update Facebook to notify the wider web), it’d have a better chance of capturing more than a couple of hours of my time a week (if that).

    Even so, my needs are different for different people; the Facebook friends lists helps a little, but only a little, in making sure that my Facebook friends knowledge about me matches my real friends knowledge about me. (That is, my life is already partitioned based on the social groups in which I reside; Facebook is making it harder for those partitions to be represented.)

  2. Chris Brogan… Thursday, January 3, 2008

    Interesting take on it, Anne. If I’ve added XYZ person as a friend on a social network, they suddenly have more of my data at their disposal. Right now, my Facebook friend adding policy is nothing. I say yes to anyone who isn’t an obvious company-pretending-to-be-auser. Otherwise, you’re in.

    Robert can take my data anywhere. I know him. It’s okay if he wants to send me a birthday card.

    BUT, if I thought robots were going to be de rigeur, which could well be, then I’m going to be more wary.

    My thoughts on this.

  3. I have not thought of it this way, but I agree, what happens on Facebook, stays on Facebook… Right…? ;)

  4. Holostatute is right on. By definition, what you are doing on Facebook is not private, esp. if you are adding someone like Scoble to your friends list (I’m assuming most people added Scoble because of his inexplicable netceleb status rather than because they genuinely know him).

    And Facebook has clearly demonstrated it could care less about your or my privacy. Its objection to scraping is almost certainly because it cannot monetize said scraping and it benefits from the lock-in it creates by forbidding it.

  5. Facebook probably has a number of objections to screen scraping, including the issue of keeping that information exclusively theirs. Still, presenting this as a situation of “it’s his data, let him do what he wants with it” doesn’t make room for the fact that it’s other people’s data, not just his.

    Once you friend someone and allow them to see various aspects of your profile you do have to understand they can do what they want with that data including moving it by hand to other places. But an unreflective “set Scoble’s data free” doesn’t recognize that there are some thorny issues about exactly whose data it is and what should be allowed with regard to it.

  6. Logical Extremes Thursday, January 3, 2008

    Both Scoble and Facebook are in the wrong.

    Scoble broke the terms of service, and potentially violated expectations of privacy of his “friends”. Facebook’s blunder, in this case, is more of a PR problem in how they handled it. On top of FB’s other recent blunders (with many issues still ongoing), it’s amazing that folks still want to play with them.

    Anne, you make excellent points. FB does need to open up, BUT also needs to provide options for the data submitters on how far they are willing to let their personal data go.

  7. Anne, there have been a lot of arguments made on this today. Yours is definitely the most reasoned and logically argued. I agree… FB clearly prevents scraping. And, while I believe that (in the absence of a scraping clause) that data that is shared with me or others, public, is in public domain, in FB’s case I agree that the users would have be treated to a fair opt-in to share the data externally.

    Jeremy Horn
    The Product Guy
    http://tpgblog.com

  8. Helzerman’s Odd Bits » Blog Archive » Alpha testing free software on someone else’s site – yep, could be a problem… Thursday, January 3, 2008

    [...] You can read reactions from Chris Brogan here and Anne Zelenka here. [...]

Comments have been disabled for this post