META Meta Platforms Inc

By Christopher Mims 

The mountain of evidence piling up this week exposes the rot at the core of Facebook Inc.

Facebook's troubling relationship with personal data, and the way that data has been repeatedly exploited, show the precarious nature of a business dependent on knowing everything it can about its own users. For a company whose very existence depends on its users turning up regularly, recent events threaten to give people a reason to reduce the amount they share on Facebook -- or leave it all together.

The latest revelations are all related to this: Facebook was, for a time, a vehicle for exfiltrating massive amounts of data about its users to developers and data miners of every stripe.

Outside developers could build games, quizzes and other apps that funneled personal information from accounts of users who willingly installed them, as well as pretty much everyone who was their friend.

Facebook allowed this data access, hoping to build a business like Apple's iPhone App Store. But the collection of Facebook data by outside developers became such a concern that Facebook eventually restricted the practice. In cases where Facebook discovered developers were using that data outside of Facebook-approved apps, the company demanded those developers delete the data.

Cambridge Analytica, the data-analysis firm that is suddenly all over the news, has worked for a number of political clients including the Trump campaign. It allegedly obtained data from the makers of one of these apps and improperly kept the data for years, despite telling the social network the records were destroyed.

"When developers create apps that ask for certain information from people, we conduct a review to identify policy violations and to assess whether the app has a legitimate use for the data," according to a statement from Justin Osofsky, Facebook's vice president of global operations. "Three years ago, we changed the product so that developers can't access the information of people's friends."

A Facebook spokeswoman says the company continues to improve its product and policies to prevent further abuse.

In the midst of this new scandal, we've been reminded that Facebook is having internal debates over how to handle revelations that Russians used the site to influence the 2016 presidential election. As the turmoil builds, politicians and regulators in the U.S. and Europe demand that Facebook make a full accounting of the abuse of its often-mysterious platform.

It won't be long before Facebook's soul-searching becomes more than an occasion to self-police and prompts users and regulators to act on their own.

A Troubling History

Again and again, we've seen two disturbing problems throughout Facebook's history. The first is that the company is unable to anticipate the ways its platform, and the incredibly powerful trove of sensitive data it produces, can be misused. In 2007, it was the way Facebook's Beacon advertising system shared users' shopping behavior and, indirectly, their life choices, with their friends and family.

Personalization in advertising is sometimes nearly indistinguishable from surveillance, and personalization is at the heart of how Facebook makes money and captures so much of the online advertising pie.

The second recurring problem with Facebook, only recently made apparent, is that the company has a powerful, often negative effect on our psychology. A variety of studies have shown that the way Facebook encourages people to passively consume friends' posts can make them unhappy. Facebook has admitted this is the case, but says it has modified its algorithm to encourage other kinds of sharing that, at least in theory, are better at positively connecting people.

Other work has shown that Facebook has the power to reinforce our biases. We think that, because our friends online espouse a view that we share, it's what the majority of the population believes. The company has contested this.

By virtue of using algorithms to target the most "engaging" content, including lucrative ads, Facebook and its ilk have become vehicles for spreading disinformation and sowing division.

Earlier this year, in an act of contrition, Facebook suggested a pivot toward individual interactions and groups would be of greater value, psychological and otherwise, to its users. But it's now apparent that even its group features are fraught with the same spammers and potential influence operations that bedeviled its news feed.

"[Facebook Groups] are how Facebook radicalizes everyday Americans," says Renee Diresta, an researcher and analyst at Data for Democracy, an independent group of data scientists. She says Facebook's algorithm for recommending groups pushes someone interested in, for example, the antivaccination movement into groups that espouse extreme political ideologies. "It's precisely because Groups facilitate trust between participants and a feeling of belonging and camaraderie that they're very powerful tools in the wrong hands," she says.

In light of these issues -- and particularly the alleged misuse of data that has pummeled Facebook's stock and reputation -- the company's options are limited. And any potential solutions could have a significant impact on the company's bottom line.

It could spend hundreds of millions of dollars to employ human moderators to police potential abuse and misuse. It could hand over its data to outside researchers, who could independently study the impact on society. It could overhaul its data strategy to radically shrink the amount of data it gathers and stores -- and monetizes.

And if it doesn't fix itself quickly, Facebook could face intrusive regulation, and even antitrust litigation.

(END) Dow Jones Newswires

March 20, 2018 17:20 ET (21:20 GMT)

Copyright (c) 2018 Dow Jones & Company, Inc.