YouTube to "anonymize" user data in Viacom lawsuit

A glimmer of good news in the wake of the uproar over the news from earlier this month that a judge had ordered YouTube to provide Viacom with the complete history of every video ever viewed on the service, along with the identity of every user who had done the viewing.

YouTube owner Google announced today that Viacom is backing down from its original demand, and that YouTube will not be providing detailed user histories to Viacom. To be more specific than what YouTube states in this blog posting, the courts have ruled that YouTube will still have to provide a record of every video watched on the service [PDF link], but it can now be delivered in anonymized form.

I'm happier, but still not totally thrilled about the ruling: You might recall that AOL's search history data was "anonymized" when it was released on the Internet in August 2006. It didn't take long for some intrepid reporters to figure out the identities of some of the actual users based on the content of the searches: If you're searching for the value of your home (and typing in your address) or doing a "vanity search" by typing in your own name, "anonymity" suddenly becomes less certain.

The problem of anonymizing history is even greater on YouTube, where I imagine many users spend copious amounts of time watching the videos they uploaded themselves. I know that my own videos on YouTube (mostly clips of my kids) have few views from the YouTube community as a whole, so it would likely take only basic investigative skills to figure out my identity based on a list of the videos I've watched. A more determined investigator will have little trouble putting a face to a name if they really want to figure out who watched their content on YouTube at any particular moment in time.

And really, what is the value of this data to Viacom anyway? Viacom is suing YouTube over copies of its TV programming being posted on the web service... but what does a list of who watched said programming gain them? I understand it might be important to the company's case to show that its content has been viewed millions of times in the aggregate, but that information is already public, as each video clip (at least, before it's removed) includes the number of times it's been watched. The only value to having that much detail, "anonymous" or not, seems threatening to users, at best.