Opinion: Twitter's privacy blunder

Twitter's decision to give companies instant access to every public post means that users' comments will be tracked, mined, and analyzed more than ever before. Perhaps it's time to think twice before you tweet.

Privacy

If you want marketers to listen even more intently to what you’re saying online, you’ll be happy to know that Twitter announced that it's giving companies a way to instantly access ever public tweet in the history of the site. According to Wired, that's more than 500 billion tweets.

If you care about privacy, you’ll be troubled by the deepening commodification of our online conversations.

Twitter might be surprised to learn that anyone’s unhappy with its initiative. After all, it's giving brands access to only public tweets, and we’ve been lulled into thinking that we check our privacy interests at the door the moment we disclose anything in public. But as the much-repeated saying goes, the “Internet never forgets,” and this means what’s said today easily can be data mined tomorrow.

Thankfully, Twitter offers meaningful privacy safeguards. It allows users to protect accounts so that anyone who wants access to a profile must be manually approved. So, tweets from protected accounts do not show up in public timelines. If you don’t take this option and instead choose an account that will display your tweets to anyone, Twitter can wash its hands of further responsibility. The common narrative is that publicity seekers only have themselves to blame if uninvited guests, including corporations, listen in on their conversations. 

At the same time, Twitter’s decision profoundly impacts obscurity: It’s now much easier for corporations to access and analyze personal information we didn’t intend for them to have. Ongoing concern exists about how journalists incorporate tweets into their stories, and now tweets are going to start showing up in Google searches. With its decision in 2010 to let the Library of Congress archive public tweets, Twitter is on a troubling course. 

Of course, the Library of Congress is a cultural institution with scholarly interests. It’s not a for-profit-corporation that wants to sell us things or sell “us” to companies. And, like many cultural institutions, the Library of Congress is plagued by funding and efficiency problems that limit what it can do with social media disclosures. As Politico recently reported, the “project is in limbo—with no end in sight.”

But the new obscurity shift poses a problem for Twitter users who find themselves unsure of how to weigh the benefits of protecting their accounts against the inconvenience that comes with it. Of course we should expect users to have a basic understanding of the tradeoff between personal data and free apps. This is 2015, after all.

But it's one thing to grasp this intellectually, and quite another thing to keep it in mind when actually using a social media platform. In fact, social media sites are designed to make you forget about the costs of disclosure and share at full speed. Friction and caution might hamper the data flow. Instant feedback and effortless socialization maximize disclosures. When you're on Twitter, your attention is drawn to things you say, replies people make to your comments and questions, and a range of remarks – jokes, confessions, astute comments – that come from folks you choose to follow.

Twitter isn't just an environment where ideas are shared. It's a highly constructed environment that's designed to make you feel like intimate exchanges are happening without giving a second thought to how everything will be monetized.

On a personal level, one of us has a protected account. But even as a privacy professional, this decision is criticized as counterproductive to the medium.

There’s a cost to protecting a Twitter account. While protected accounts are obscurity-friendly, they’re also harder to follow. And they’re harder for others to interact with.

Explicit permission is needed to follow a protected account. By contrast, public accounts can be followed automatically.

Trying to decide whether to follow a protected account? Good luck. All of the tweets and the list of followers are locked down. All you’ve got to go in is the account name – and the name doesn’t necessarily clue anyone in as to who runs the account.

Retweeting also is comparatively labor intensive on protected accounts. Twitter’s system is designed so that only public tweets can be widely shared at the touch of a button. While this may seem like a trivial point of friction, it can deeply impact whether a dominant form of communication on the platform gets exercised.  

Bottom line: How can anyone reasonably determine if Twitter’s transaction costs are worth taking on without having a clear sense of what’s at stake if you brush them off?

Joseph Jerome, Policy Council at the Future of Privacy Forum (where one of us is a senior fellow), sees Twitter’s approach to cozying up with companies as more complicated than meets the eye. He argues Twitter should be more sensitive to how users view the platform, and it should provide more choices than currently are available.

“While Twitter has been clear that everything you post is public, Twitter must recognize that tweeting publicly and selling a fire hose of data that companies can then link to emails and actual customer identities are not the same thing, and begins to stress user expectations. Twitter should provide an opt-out to their users for this sort of practice.

Also, companies should realize that even though their customers may have posted something 'publicly' on Twitter, customers are likely to be quite surprised that every burp and belch they make on Twitter is being used to profile and market to them. Companies that dip into this data will need to think hard about which uses are appropriate and which might be deemed wildly out of context.”

It’s not clear exactly how much of Twitter’s treasure trove it will reveal to businesses and we also don’t know the manner in which this information will be shared.

Will Twitter follow Facebook and create a privacy-protective interface for advertisers that obscures the identity of its users? Or does the "public" nature of the tweets mean Twitter doesn’t feel compelled to protect the privacy of its users in giving businesses exclusive backstage access?

Twitter cleverly used the concept of “public” to leverage our intuitions about information that is “not private,” but the reality of public data is hardly that simple. Both policy and social norms are increasingly aligned against the idea that mere accessibility and localized disclosure strip data subjects of any privacy interest in that information and absolve companies from any privacy obligations.

When talking to our friends and colleagues on social media, it’s easy to forget we’re really speaking directly to companies that need to monetize our data to grow. If these companies don’t give us good options for responding to diminished obscurity, they aren’t taking our privacy seriously.

Evan Selinger is an associate professor of philosophy at Rochester Institute of Technology. Follow him on Twitter @EvanSelinger. 

Woodrow Hartzog is an associate professor at Samford University’s Cumberland School of Law. Follow him on Twitter @Hartzog.