Using data gathered from nearly 14,000 Twitter users, the team of scientists analyzed information about the content of people’s tweets. They found that using information from the tweets of a few of someone’s contacts — just eight or nine friends — made it possible to predict that person’s tweets as accurately as using data from their own Twitter feed. This means that it’s possible to predict your Twitter content from seeing your friends’ tweets, even without having access to your Twitter.
Even more concerningly, the behavior of users who have left Twitter or who never joined in the first place can be predicted in the same way. If your friends are on Twitter, it is possible to use their data to provide 95 percent “potential predictive accuracy” of how you would behave on the platform in the future.
Mathematician James Bagrow, lead author of the paper, points out that this means that signing up for a social network carries more responsibility than we’re generally aware of: when you sign up for a platform like Facebook or Twitter “you think you’re giving up your information, but you’re giving up your friends’ information too!”
This research questions the assumption that people make individual choices about privacy matters. The traditional view is that users knowingly give up some privacy over their data in return for free use of a service that they find beneficial. But this shows that privacy is a bigger issue and that individuals cannot personally control the way that their information is spread through social media. Even if you never join a social media site it is still possible to profile you through your friends and for companies or governments to get access to information about your political affiliations, religious beliefs, shopping habits, and so on.
“There’s no place to hide in a social network,” says Lewis Mitchell, co-author of the study. Bagrow agrees: “You alone don’t control your privacy on social media platforms. Your friends have a say too.”