Exclusive: Facebook will ban misinformation on the vote in the upcoming US elections

MENLO PARK, California (Reuters) – Facebook Inc (FB.O) will prohibit misinformation about voting conditions and factually verify false reports of violence or long queues at polling stations prior to mid-term elections in the United States of the month next, said company officials in Reuters, the latest effort to reduce the number of voters handled.

The largest online social network in the world, with 1.5 billion daily users, has stopped banning all fake or misleading messages, which Facebook has leaked because it would probably increase its expenses and leave the door open to censure charges.

The latest initiative concerns a sensitive area for society, which has been criticized for its lax approach to misinformation and misinformation campaigns, which have apparently affected the outcome of the presidential election. of 2016, won by Donald Trump.

The new policy was revealed by Nathaniel Gleicher, head of Facebook's cyber security policy, and other senior executives of the company.

The ban on misinformation about voting methods, to be announced Monday, comes six weeks after Senator Ron Wyden asked Chief Operating Officer Sheryl Sandberg how Facebook will counter messages to remove votes. for example by indicating to certain users that they could vote. text, a hoax that has been used to reduce participation in the past.

Information on voting methods is becoming one of the few areas in which lies are banned on Facebook, a policy imposed by what the company calls moderators "community standards", although the application of its standards has been uneven. This will not stop the vast majority of misleading messages about candidates or other electoral issues.

"We do not think that we should remove from Facebook the elements shared by genuine people if they do not meet the standards of this community, even if they are false," said Tessa Lyons, Product Manager for Thread. Facebook news, which shows what users have to offer. friends share.

The links to discouraging reports on polling sites that might be inflated or misleading will be referred to the fact-checkers as part of the new policy, Facebook said. If then marked as fake, the reports will not be deleted but will be seen by fewer friends of the poster.

Such partial measures leave Facebook more prone to manipulation by users seeking to affect the election, according to critics. Russia, and potentially other foreign parties, are already making "intrusive" efforts to interfere in the upcoming US elections, said the head of Trump's national security team in early August.

A few days earlier, Facebook had announced the discovery of a coordinated political influence campaign aimed at misleading its users and sowing discord among voters, removing 32 pages and accounts from Facebook and Instagram. Congressmen informed by Facebook said the methodology suggested Russia's participation.

Trump has disputed claims that Russia has attempted to interfere in the US elections. Russian President Vladimir Putin denied it.

PROHIBITION OF WEIGHING THE MATERIALS

Facebook has introduced a worldwide ban on false information about when and where to vote in 2016, but Monday's decision goes further, including publishing articles about exaggerated identification requirements.

Facebook executives are also wondering if they should follow Twitter Inc. (TWTR.N) a recent policy change to ban publications linking to pirated content, Gleicher told Reuters in an interview.

The broadcast of hacked emails from Democratic party officials probably played a role in overturning the 2016 presidential election in Trump, and the national intelligence director, Dan Coats, warned that Russia was recently trying to hack and to steal information from US candidates and government officials.

However, a blanket ban on pirated content would limit exposure to other media that, according to some, would serve the public interest, such as the "Panama Papers", which have made public information available. widespread use of foreign tax havens by the world's rich.

A few months ago, Facebook executives briefly discussed the ban on all political ads, which generate less than 5% of the company's revenue, sources said. The company rejected this idea, as product managers were reluctant to leave advertising money on the table and politicians argued that blocking political ads would favor incumbents and the rich. activists who can better afford to switch to television and print ads.

Instead, the company checks the proof of national residence of political ad buyers and keeps a public record of who bought what.

Facebook also questions the authenticity of personal accounts. It can use an automated activity detected to disable propaganda-propagating pages, as it did last week, but it does not require a phone number or other individual proof of identity before allowing users to open accounts.

With regard to false information, Facebook has suspended the total ban, thus limiting the circulation of articles marked as false by verified fact checkers. However, this approach can leave investigators overwhelmed and able to tackle only the most virulent hoaxes.

"Without a clear and transparent policy to curb the deliberate spread of false information that applies across platforms, we will continue to be vulnerable," said Graham Brookie, head of the Atlantic Council's Digital Forensic Research Lab.

(This version of the story has been corrected to indicate that misinformation about polling stations will be checked, not that it will be totally prohibited in the first paragraph.) In the eighth chart, an explanation of how the reports on the polling stations will be processed will be provided.)