Skip to Content

Facebook’s Rules for Experimenting on You

The social network has been slammed for research that meddled with its users but says new rules will keep experimentation ethical.
June 15, 2016

Questionable research ethics have gotten Facebook in trouble in recent years, but the company says it now has a system in place to prevent that from happening again.

Before they can use information about you in science projects, Facebook’s data scientists must run their research proposals through a formal review process to make sure the research is ethical and sound in the areas of privacy and security. Some details of the process, which the company says has been in place for two years, were revealed in a new paper published in the Washington and Lee Law Review.

The research practices of companies that collect and analyze enormous amounts of data about their users—Facebook being the most prominent—represent a growing area of concern to policymakers. Last month the White House called on users of “big data” to create better systems for tackling challenges related to data ethics, security, and privacy.

Two years ago Facebook drew heavy criticism after publishing a research paper describing an experiment in which company scientists manipulated the news feeds of users to test whether they could influence their emotions. It has faced a backlash from academics critical of the methods used in other studies as well.

The review process described in the new paper is in large part a “response to feedback we’ve received,” Molly Jackman, Facebook’s public policy research manager, said Tuesday at a meeting about big data research ethics at the Future of Privacy Forum in Washington, D.C.

The crux of it is a standing committee of five employees, including experts in law, ethics, communications, and policy. This is inspired in part by the federal model for institutional review boards, which vet academic research proposals and identify ethical concerns.

Much of the research that Facebook does is the type that is generally exempt from an academic institutional review board, said Jackman. Other projects that involve more sensitive subjects or populations warrant further review. The suicide prevention program the company launched Tuesday is an example of a project that falls in that category, she said. Senior managers decide whether proposals should either be expedited or reviewed by the committee.

Jackman said the process is vital because the company’s vast data about human behavior puts it in a position to “push on the boundaries of science.”

But while Facebook’s data may contain insights that could help society, company-led research inevitably raises tricky questions. For instance, how will the members of the review board separate the company’s interests from the interests of the research subjects or the greater public? “You need someone on that review board who can’t get fired,” Rey Junco, a professor of education and human computer interaction at Iowa State University, said at the Future of Privacy Forum meeting.

Another controversial issue is consent. Whereas subjects taking part in federally funded research must provide “informed consent” before they can be included, Facebook users consent to research when they sign up for the service. Junco said researchers should place a greater emphasis on consent, so that users have the opportunity to “self-determine what they are going to be engaging in.”

Keep Reading

Most Popular

A view of clouds illuminated by sunlight
A view of clouds illuminated by sunlight

We can’t afford to stop solar geoengineering research

It is the wrong time to take this strategy for combating climate change off the table.

conceptual illustration showing various women's faces being scanned
conceptual illustration showing various women's faces being scanned

A horrifying new AI app swaps women into porn videos with a click

Deepfake researchers have long feared the day this would arrive.

Death and Jeff Bezos
Death and Jeff Bezos

Meet Altos Labs, Silicon Valley’s latest wild bet on living forever

Funders of a deep-pocketed new "rejuvenation" startup are said to include Jeff Bezos and Yuri Milner.

ai learning to multitask concept
ai learning to multitask concept

Meta’s new learning algorithm can teach AI to multi-task

The single technique for teaching neural networks multiple skills is a step towards general-purpose AI.

Stay connected

Illustration by Rose WongIllustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.