Propaganda examples in social media

The Internet Research Agency's operators ran hundreds of American personas, each with a name, a location, a political identity, and a following, some of which reached hundreds of thousands of followers and were quoted by mainstream outlets as real citizens. The personas argued with each other across the political spectrum to deepen division. Court filings and the Senate report reproduce their character notes.

"Oliver Taylor," a supposed University of Birmingham student and freelance writer, placed opinion pieces in the Jerusalem Post, the Times of Israel, and other outlets in 2019 and 2020 attacking activists. A Reuters investigation found no such student, a profile photograph that experts identified as generated, and a persona that vanished when exposed. The editors had never spoken to him.

When Gabon's government released a New Year's address by President Ali Bongo after months of his absence for medical treatment, opponents declared the video a deepfake, and the claim became part of the justification for an attempted coup days later. Analysts found no evidence of forgery. The episode is the standard case of the liar's dividend: the possibility of a fake did the political work.

During the 2016 U.S. election, sites such as "abcnews.com.co" and the "Denver Guardian," the latter a nonexistent newspaper whose fabricated story about an FBI agent's death was shared more than half a million times, imitated real outlets' names and designs. The episode made the imposter site a recognized category and prompted platforms to label domains.

Ferrara's analysis of some twenty million election-related tweets estimated that roughly a fifth of the accounts involved were bots, generating a larger share of the volume, and that the automated accounts were retweeted by humans at similar rates to human ones. The finding, that people could not tell, is the device's whole premise.

The Gamergate campaign coordinated in forums and chat channels to flood the accounts of women game developers and critics with abuse, to mass-report their posts, and to pressure advertisers. Massanari's analysis of the platform dynamics described how a site's design, upvotes, anonymity, and weak moderation, made it "a fertile ground" for the device.

McDonald's promoted #McDStories in January 2012 to collect happy customer anecdotes, and within hours the tag filled with accounts of food poisoning, labor complaints, and jokes; the company withdrew the promotion the same day. The episode named the corporate form, "bashtagging," and became the case every social-media guide cites.

Working Families for Wal-Mart was launched in 2005 as a grassroots group of supporters and was revealed to be organized and funded by the company through its public-relations firm Edelman, which also ran a blog by a couple "traveling America in an RV" who turned out to be on the firm's payroll. Edelman apologized in 2006.

Every major storm since Hurricane Sandy in 2012 has been accompanied by the same recirculated images: a shark on a flooded highway, a dramatic cloud over a city that was not the one named, footage from a previous year's flood. Fact-checkers documented the same photographs reappearing for Harvey in 2017 and Florence in 2018 with new captions each time.

Paris and Donovan's report for Data & Society cataloged the techniques, recontextualising, speeding and slowing, lookalikes, face-swapping with consumer apps, and documented cases in which each had circulated as genuine, from doctored footage of public figures to misattributed war footage. They argued that the cheap end of the spectrum, not the expensive one, is where policy attention belonged.

Internal documents disclosed by the former Facebook employee Frances Haugen in 2021 showed that the company's 2018 ranking change, which weighted comments and reactions more heavily, had increased the reach of divisive and angry content, that the company's researchers had documented the effect, and that publishers and political parties had told the company they were producing angrier content to keep up.

The site Upworthy grew to tens of millions of monthly readers on a headline formula, "curiosity gap" headlines tested twenty-five variants at a time, that was copied across the industry. When Facebook adjusted its ranking in 2014 to penalise the pattern, the site's traffic fell by more than half, which measured how much of its audience the device had supplied.

Reporting in 2018 showed that Cambridge Analytica had obtained data on tens of millions of Facebook users through a personality-quiz app and offered campaigns targeting by inferred psychological profile. Whatever the actual efficacy, the disclosures led to parliamentary and congressional hearings, a record fine for Facebook, and platform ad libraries that made dark posts partly visible.

Online communities produced and coordinated political memes during the 2016 campaign at a scale that reached mainstream coverage, and the Internet Research Agency's accounts, the Senate report found, relied heavily on the format, posting image macros on race, religion, guns, and immigration to millions of followers. The Anti-Defamation League's 2016 designation of a cartoon frog as a hate symbol, in some contexts, marked the medium's arrival as propaganda.

After MH17 was shot down over eastern Ukraine in July 2014, Russian state media and officials advanced a series of incompatible accounts: a Ukrainian fighter jet, a Ukrainian missile aimed at another aircraft, a CIA plot, a plane pre-loaded with corpses. None was withdrawn as the others appeared. The Dutch-led investigation's finding that a Russian-supplied Buk missile was responsible met the same treatment.

David Malki's Wondermark strip "The Terrible Sea Lion" showed a couple pursued by a sea lion demanding, with impeccable courtesy, that they justify a remark one of them had made about sea lions. The strip was adopted within weeks as the name for the behavior it drew, and the word appeared in the Oxford English Dictionary's watch list soon after.