Facebook’s data collection divides us more than it connects us

41
Naomi Field

It’s a typical day in 2018 America. The sun is shining, summer is around the corner and another scandal’s just broken out about election interference via personalized propaganda on social media. Oh, to be young and psychologically manipulated by forces beyond our control!

If you don’t know what I’m talking about, here’s the scoop: Cambridge Analytica is a voter profiling company with ties to Donald Trump’s 2016 election campaign. The company stole data from over 87 million people’s Facebook profiles using an app created by researcher Aleksandr Kogan; only 270,000 people had consented to the company harvesting their data. The app scraped personal data from both the profiles of the users who downloaded it and from their friends’ profiles. The company then used the data to target Facebook users with highly personalized ads for Trump’s campaign.

A similar scandal recently occurred on Tumblr. According to New York Magazine, the website’s staff discovered that 84 different Tumblr accounts were linked to the Internet Research Agency (IRA), a Russian “troll farm” that aimed “to shift the election in Trump’s favor … by stoking social tensions.” The troll blogs were run by real people, not bots, and blended into Tumblr’s user base by posting memes and other innocuous content amongst more blatantly political posts.

On the surface, the scandals are similar: both used propaganda to convince users to vote for Trump, and both involved an unnervingly personal element (either manipulating users’ personal interests or posing as the friendly meme blog next door). But the Facebook scandal is more sinister because of the difference in the websites’ natures. While Tumblr allows users to share as much or as little personal information as they’d like, Facebook requires and then categorizes its users’ information — from their birthdays to their interests to their political views. That may allow for users to get to know each other better, but it also makes them susceptible to dangerous manipulation — from Russian hackers as well as other criminals like identity thieves — and ultimately causes more division than connection.

To understand the difference between the two situations, look at the Tumblr staff’s response post to the IRA scandal. They claimed to have emailed every user who “reblogged, replied to or followed an IRA-linked account” to notify them of the issue. The post also linked to a public record of IRA-linked usernames, warned against divisive rhetoric, emphasized the importance of fact-checking and urged users to vote.

The response is straightforward and gives Tumblr users the information they need to avoid posts with divisive or false content. But the stakes are lower on Tumblr; reblogging someone else’s false post is a far less serious breach of trust than having your personal data stolen.

In Facebook CEO Mark Zuckerberg’s response to the Facebook scandal, he claimed that Facebook will also be taking more security precautions in the future, including restricting access to data and testing out a new tool for ad transparency so that users will know who’s paying for the ads they see.

“It’s not enough to just connect people. We have to make sure those connections are positive and that they’re bringing people together,” Zuckerberg said.

The response appears to be a direct reference to Facebook’s new mission statement, which the company changed June 2017. Previously, Facebook’s goal was to “[make] the world more open and connected;” now, it’s to “give people the power to build community and bring the world closer together.” In Zuckerberg’s explanation for the change, he claimed that Facebook would meet this goal by introducing new AI tools that would recommend “Groups” for users based on personal information such as their location and gender.

Sound familiar? It should. According to Vox, Cambridge Analytica used data such as users’ Facebook “likes” to determine facets of their personality (examining over 300 of a user’s “likes,” for example, could supposedly tell the company more than “what a person thought they knew about themselves”) and then showed them targeted ads based on that information. Bots then redistributed the most popular ads depending “on where they were popular and who they appealed to.” The data even informed Trump’s travel plans and speech topics.

So, although Facebook collected and categorized personal data under the guise of creating positive connections, it got the opposite result; the same user data that Facebook intended for harmless community-building allowed for the rapid spread of propaganda.

Hopefully, the security features that Zuckerberg plans to implement will be more effective in the future, since companies like Facebook should ultimately be responsible for their users’ safety. If they’re not, though, there’s not much we can do beyond attempting to delete as much information as possible.

Be mindful not only of what information you share on the Internet but how you share it. Social media sites like Tumblr and Twitter, which don’t request neatly categorized personal information, may not be entirely safe from manipulation — the Tumblr scandal proves that. But they are safer, and right now, that may be the best we can do.

Natalie Ray is a sophomore English major. She can be reached at nray@oxy.edu.