Facebook to Start Fact-Checking Photos, Videos
Facebook to Start Fact-Checking Photos, Videos
The social-media company will use technology and human
reviewers to help flag false content
By Micah Maidenberg Updated Sept. 13, 2018 4:06 p.m. ET
Facebook Inc. will begin fact-checking photographs and
videos posted on the social media platform, seeking to close a gap that allowed
Russian propagandists to promote false news during the last U.S. presidential
election.
The company said Thursday it will use technology and
human reviewers to try to staunch what it called in a statement “misinformation
in these new visual formats.” Previously, the company’s efforts had been
focused on rooting out false articles and links.
“The same false claim can appear as an article headline,
as text over a photo or as audio in the background of a video,” Facebook
product manager Tessa Lyons said in the statement. “In order to fight
misinformation, we have to be able to fact-check it across all of these
different content types.”
During the 2016 presidential election, a Russian group
called the internet Research Agency helped its workers create graphics and
videos that could spread misinformation via social media networks, according to
special counsel Robert Mueller’s indictment earlier this year of three Russian
companies and 13 citizens of the country.
A Russian-linked group, for example, once altered a photograph
of a woman carrying a sign at a pro-immigration rally in Arkansas. The original
text on her sign said, “no human being is illegal” and the doctored photo said
“give me more free shit.” It was posted to Facebook in August 2017, where it
was liked or shared hundreds of times.
Yet, as with other technology companies, Facebook will
face a significant challenge in designing algorithms that are able to catch
doctored photographs and videos, or those that have been posted without
context. A Facebook spokeswoman said the company’s efforts to fact-check video
and photos will rely on technology but also human reviewers who work for groups
certified by an organization called the International Fact-Checking Network.
The company is targeting video and photo content that has
been “manipulated, taken out of context, or includes a false text or audio
claim,” Ms. Lyons said in Facebook’s statement.
She acknowledged that “figuring out whether a manipulated
photo or video is actually a piece of misinformation is more complicated; just
because something is manipulated doesn’t mean it’s bad.” Ms. Lyons added the
company can use technology “to identify different types of manipulations in
photos, which can be a helpful signal that maybe something is worth having fact-checkers
take a look at.”
In addition to technology, Facebook will rely on user
feedback to help flag false content in videos and graphics, similar to what it
does now with articles. The firm also examines comments on a post that might
indicate misinformation and whether those sharing content have a history of
sharing items rated false by fact-checkers.
Comments
Post a Comment