MSFT Microsoft Corporation

By Sarah E. Needleman 

False and misleading information is spreading in new ways across social media, as people seeking to influence political campaigns or achieve illicit goals attempt to stay ahead of platform operators trying to stop them.

The new methods range from employing real people -- instead of relying solely on bots -- to post messages that appear authentic and trustworthy, to sharing fabricated images, audio and video, researchers say.

Purveyors of disinformation are experimenting with these techniques largely because Twitter Inc., Facebook Inc. and other social-media companies have beefed up efforts to squash their campaigns, though some critics want to see the companies do more.

The companies began sharpening their focus on coordinated campaigns intended to sway public opinion after the identification of Russian government interference in the 2016 U.S. presidential election. A report from a U.S. Senate committee released earlier this month criticized U.S. tech giants for helping spread disinformation during the 2016 campaign and called for better coordination of efforts to prevent similar activity for elections in 2020.

"People think it's a problem that can be fixed," says Nathaniel Gleicher, head of cybersecurity policy at Facebook. "But we know these actors are going to keep trying. That's the one guarantee we have."

Lee Foster, an analyst at cybersecurity firm FireEye Inc., says manipulators are also changing their tactics because the public has become more aware that disinformation operations are carried out on social media. Mr. Foster was part of an investigation into a coordinated disinformation campaign on Facebook, Twitter and other large social-media platforms last year believed to promote Iranian interests.

A veneer of truth

Interest groups, governments, companies and individuals are motivated for several reasons to spread disinformation, according to researchers. Some are looking to promote a political agenda or ideological position, while others want to draw social-media users to a website so they can generate advertising or other revenue. In some cases, the messages dispersed are merely pranks carried out for entertainment purposes.

The results of these efforts can be significant, with the potential to distort elections, divert attention from other societal issues or instill fear. They can also be deadly. Last year, more than 20 people were killed in India by mob violence after false rumors about child-trafficking activity spread on Facebook's WhatsApp messaging service.

Increasingly, people seeking to spread disinformation are becoming less reliant on bots to flood platforms with messages, having humans behind keyboards do the work instead, researchers say. They say organizers are paying people who have large online followings to amplify their messages without disclosing the transactions.

Campaign organizers are also taking over dormant accounts under new guises and encouraging people to post messages in support of their cause, researchers say. These organizers, in many cases, obscure their identities and intentions and can dupe users into spreading false information through their networks.

"You have less bot activity and now people publishing multiple times a day," says Claire Wardle, director of First Draft, a nonprofit that supports research into disinformation campaigns on the internet and is funded in part by Alphabet Inc.'s Google, as well as Facebook and Twitter. "They're using other people already very passionate about certain issues to push weaponized content."

For example, ahead of the 2017 French presidential election, a group of Marine Le Pen supporters sought to promote her campaign by simultaneously posting tweets with the same hashtag and had bots do the same, according to Ben Nimmo, head of investigations at New York-based Graphika Inc., a social-media-analytics firm. With this approach, people are looking to give the false impression that their posts have gone viral organically, which can help them in turn attract coverage by legitimate news outlets, he says.

Hashtags such as #LaFranceVoteMarine and #Marine2017 began trending on Twitter across France, raising Ms. Le Pen's profile by gaining coverage from French media that portrayed her supporters as powerful and effective online.

"It was a lot more sophisticated than just turning on a bot army," Mr. Nimmo says.

Also, rather than solely pumping out falsehoods, organizers are mixing truth with deceptive content on the larger social platforms -- or with inflammatory content on smaller ones that have looser content controls such as Gab and 4chan -- to gradually build up followers and make themselves appear more credible, researchers say. For instance, a real photo may be posted on these smaller sites with a highly sensationalized or misleading caption to inflame social divisions.

"Gab allows all protected political speech, no more, no less," said Andrew Torba, chief executive of the social network, in a statement. Representatives for 4chan didn't respond to a request for comment.

Fake images and audio

So-called deep fakes -- phony images, video or audio created with artificial intelligence -- are another new tool used to deceive social-media users, researchers say. These differ from already common "cheap fakes" -- deceptive alterations of real content, such as the doctored video of House Speaker Nancy Pelosi (D., Calif.) that went viral on YouTube and Facebook in May.

"Deep fakes are worrisome because they are very hard to detect by the lay public and journalists alike," says Joan Donovan, director of research at Harvard University's Shorenstein Center on Media, Politics and Public Policy. "Looking forward to 2020, these newer tactics are going to be difficult to manage and defend against."

Another concern, says FireEye's Mr. Foster, is that people will claim video or audio of them acting badly is a deep fake, even when that isn't the case. "Their mere existence will provide plausible deniability for the actions or comments that real people do make," he says.

Such new techniques are coming into play as coordinated attacks are expanding geographically, originating from 70 countries this year as of September, up from 48 in 2018 and 28 in 2017, according to a study by researchers at Oxford University. The findings also show that 47 countries have used state-sponsored groups to attack political opponents or activists so far this year, nearly double the number from last year.

Platforms fight back

Facing mounting pressure from governments and public outcry, social-media companies have worked to thwart platform manipulation. Twitter and Facebook, for example, have suspended tens of millions of accounts affiliated with disinformation campaigns across the globe, while Reddit Inc. this year said it removed more than 10.6 million accounts with compromised login credentials.

The companies have also ramped up hiring of staff dedicated to spotting coordinated attacks. Facebook said it now has 30,000 employees focused on user safety and security, three times as many as it had 18 months ago. But liberal- and conservative-leaning politicians and interest groups as well as free-speech advocates have panned social-media companies in recent years for either dragging their feet on content-moderation controls or going too far to alter genuine discourse. In a speech at Georgetown University earlier this month, Facebook Chief Executive Mark Zuckerberg said the company must continue to stand for free expression even though it has worked to remediate concerns about misinformation and hate speech on the social network.

Social-media companies also have been sharing data they have collected on coordinated attacks with researchers. Some companies have also created deep-fake videos to help researchers develop detection methods. Facebook earlier this year teamed up with several partners including Microsoft Corp. to launch a contest to better detect deep fakes. Mr. Gleicher of Facebook says the spread of false and misleading information on social media is too difficult for any one company to tackle alone.

That said, companies need to be cautious when they share information about coordinated attacks with the public to avoid giving people insight into how to game the system in the future, says Yoel Roth, head of site integrity at Twitter.

Social-media firms also don't want to inadvertently inflate the magnitude of manipulation campaigns by disclosing too many details about how they occur, Mr. Gleicher says. "Sophisticated actors try to make themselves look more powerful than they are," he says.

Ms. Needleman is a Wall Street Journal reporter in New York. She can be reached at sarah.needleman@wsj.com .

(END) Dow Jones Newswires

October 24, 2019 19:17 ET (23:17 GMT)

Copyright (c) 2019 Dow Jones & Company, Inc.