Facebook Reports Gains in Removing Objectionable Content
November 13 2019 - 7:24PM
Dow Jones News
By Jeff Horwitz
Facebook Inc. reported gains in detecting hate speech,
child-abuse imagery and terrorist propaganda on its platforms,
arguing that its willingness to publish statistics on the removal
of objectionable content shows its commitment to transparency.
"The systems we've built for addressing these issues are more
advanced than what any other company has," Chief Executive Mark
Zuckerberg said Wednesday, adding that other internet companies
have avoided making similar disclosures because "they don't want to
admit they have a problem too."
The Facebook co-founder was unusually blunt in his criticism of
other internet and social-media platforms on a call with reporters,
but he didn't specifically name any competitors nor say where their
detection efforts and disclosure fall short.
Facebook, Alphabet Inc.'s Google and Twitter Inc. are among
large social-media companies that release reports aimed at giving
the public a window into their businesses. These firms have
responded to persistent complaints about illegal and offensive
content on their platforms by providing metrics to judge their
progress in removing it, sometimes before it gets flagged by
users.
Facebook's enforcement report shows the volume of problematic
content it hosts remains staggering, with more than 12 million
pieces of child nudity or sexual-abuse content alone being removed
from the main Facebook platform and Instagram during the September
quarter. But the company said the actual frequency with which such
material is viewed is so small that it can't be reliably measured,
with such posts accounting for less than 0.04% of what users
actually saw.
The report marked the first time the company reported
content-enforcement statistics for Instagram, which is
significantly less successful than the Facebook platform at
detecting terrorist propaganda, child exploitation and self-harm
content before it is reported by users.
On Facebook, all but 3% of suicide and self-harm content is
removed from the platform without reports from users, the company
said. Instagram users are still responsible for flagging 21% of
similar content that is eventually removed.
Vishal Shah, Instagram's head of product, said the disparity
partially reflects Facebook having more mature artificial
intelligence systems, and he pledged Instagram would make gains
over time.
Facebook also touted its progress this year in removing posts
that violate its policy barring illicit firearm and drug sales. It
said about 6.7 million pieces of such content was removed from
Facebook in the most recent quarter, up from nearly 1.5 million in
the first quarter.
Facebook has faced criticism over its Marketplace operations in
connection with gun sellers being able to dodge prohibitions
designed to end such activity.
U.S. Attorney General William Barr and international officials
in recent months have criticized Facebook's plan to roll out
encrypted messaging across its services, a step that would make it
harder for either the company or law enforcement to monitor for
illegal activities.
Mr. Zuckerberg acknowledged "real tension" between encryption
and safety. At the same time, he said Facebook should be credited
for helping make society safer by doing a better job of policing
the company's platforms.
Write to Jeff Horwitz at Jeff.Horwitz@wsj.com
(END) Dow Jones Newswires
November 13, 2019 19:09 ET (00:09 GMT)
Copyright (c) 2019 Dow Jones & Company, Inc.
Twitter (NYSE:TWTR)
Historical Stock Chart
From Aug 2024 to Sep 2024
Twitter (NYSE:TWTR)
Historical Stock Chart
From Sep 2023 to Sep 2024