SXSW: Is Privacy on the Social Web a Technical Problem?

How to deal with user privacy on social networks as they grow, mature and become more sophisticated has been a frequent topic of conversation at this year’s SXSW — and not just in researcher Danah Boyd’s keynote address that argued aggregating public information can be a privacy breach, and slammed Google (s GOOG) and Facebook for their missteps with users’ expectations.
Is privacy just a technical problem? That’s what Google engineer Brett Slatkin, co-creator of the PubSubHubbub real-time syndication protocol, proposed on a Saturday morning panel. WebFinger, a cross-platform standard that conveys explicit privacy settings Update: user preferences, which could include explicit privacy settings from one social network to another, could take care of understanding the relationships between users and the information they want to control, Slatkin said. He added that he felt that the reason users are confused about privacy is because of inconsistency among the social sites they use.
But Microsoft (s MSFT) program manager Dare Obasanjo contended that for-profit social web companies’ interests will always be at odds with user privacy, because there’s too much value in harnessing the crowd for things like Twitter’s trending topics and search. He said he felt the industry needs to “clean up [its] act” on privacy, citing Netflix’s cancellation of its second Netflix Prize contest this week due to concerns that the dataset it provided competitors could be matched to its customers.
Obasanjo also argued that approaches like WebFinger might not work because asking users to specify privacy controls introduces friction into their use of your site.
Collecta co-founder Jack Moffitt disagreed with Google’s Slatkin, but for a different reason. “I don’t think the solution is purely technical,” he said. “I don’t think users will understand the repercussions of these decisions.”
Google’s approach to the social web, where it has fallen behind competitors like Facebook, is heavily focused around standards, and Slatkin described things like OStatus, DiSo, Activity Streams and AtomSource as the key to solving all sorts of usability problems, privacy included. But Slatkin is far from the only true believer in standards. Jon Phillips of, on a later panel that focused on whether tweets can be copyrighted, spoke about how interoperability can help pass along information as to who owns information and where it comes from. Both Slatkin and Phillips said they feel that if information can be properly sourced and transmitted, rather than replicated, it can be better owned by users.
Still, companies like Google and Facebook don’t have the luxury of being able to start with a clean slate of user expectations and privacy settings, because they’re evolving to adapt to the enormous growth of online sharing and the increasing influence of social sites on the rest of the web through things like real-time search. Kerfuffles like the ones Boyd highlighted in her keynote are the result of changing what information is private and public after people have come to expect something else.
In yet another discussion at the conference, Flickr (s YHOO) head of product Matthew Rothenberg said that his team has had to deal with similar privacy issues concerning displaying publicly when a user has favorited a photo on the site (which has always defaulted images to public). “Our bet is that by enforcing public behavior we’re going to change the nature of what they’re doing,” he said. But “you have to properly educate people as to what you’re doing.”
So privacy, it seems, is indeed a technical problem, but also a cultural problem, an educational problem and a business problem. Not to mention endless conversation fodder.
For the GigaOM network’s complete SXSW coverage, check out this round-up.
Updated to clarify WebFinger based on emailed comments from Slatkin.
Related content from GigaOM Pro (sub req’d):
How Facebook Should Fix Its Privacy Problem