Why Twitter and Facebook treat Bloomberg’s tricks differently

Note: This is something I originally wrote for the daily newsletter at the Columbia Journalism Review, where I’m the chief digital writer

Donald Trump’s presidential campaign in 2016 used to be the high-water mark (or low-water mark, depending on your perspective) for the aggressive use of social networks in targeting voters. This time around, it’s a different billionaire — Mike Bloomberg — who is testing the limits of what is permitted on various platforms, and so far they seem to be treating him very differently. To take just one example, Twitter recently suspended 70 accounts because they each posted identical pro-Bloomberg messages as part of the campaign’s social-media marketing blitz. According to a number of reports, Bloomberg has hired hundreds of social-media “influencers” to post messages about him on various networks, and is paying them $2,500 a month. But while Twitter reacted harshly, Facebook seems untroubled by this kind of behavior. It says such posts are fine so long as they are labeled as ads.

In one sense, this difference of opinion on the Bloomberg campaign’s digital strategy is tied to the corporate DNA of each company. Historically, Twitter always placed a premium (for better or worse) on making “authentic” communication on its platform as friction-less as possible, although it has muddied those waters somewhat by introducing images, auto-play videos, algorithmic filtering, and advertising. But regardless, from Twitter’s point of view, the posting of dozens or hundreds of identical messages — whatever their content or intention — meets the definition of spam, or what the company calls “platform manipulation,” and therefore must be removed. Twitter’s rules forbid creating multiple accounts to post “duplicative content,” posting identical or similar Tweets or hashtags from multiple accounts operated by a single individual or corporate entity, and “coordinating with or compensating others to engage in artificial engagement or amplification.”

Coordinating with and/or compensating others for generating artificial engagement and amplification, of course, is exactly what the Bloomberg campaign is designed to do. To Twitter, that might look like spam, but to Facebook it just looks like advertising, and therefore the social network is more than happy to facilitate it. In fact, as the Trump campaign discovered in 2016, the company isn’t just happy to have it exist on the platform, if you are important enough and pay Facebook enough, it will embed Facebook staffers inside your campaign and help you do it better. But isn’t that kind of thing what the company calls “coordinated inauthentic behavior?” Apparently not. Facebook says that term is reserved for coordinated campaigns where people pretend to be other people, not campaigns where people pretend to like or admire someone they don’t. That’s just called advertising.

In a way, what Twitter and Facebook’s polar opposite approaches illustrate is just how muddy the media landscape has become thanks to social networks like theirs, and especially the difference between advertising and non-advertising. This became obvious when Facebook (and to a lesser extent, Twitter and YouTube) was called before Congress to answer for the way that Russian trolls used the platform to distribute misinformation about the 2016 election. The Congressional inquiry focused on ads that were bought and placed on Facebook by various foreign agents, including Russia’s Internet Research Agency — but this focus mostly missed the point of a network like Facebook, which is that in a very real way, everything posted on the service is essentially an ad, or at least can be made to behave like one by paying the company to “boost” the post with its News Feed algorithm.

This confusion even extends to the way Facebook treats “influencer” ads: The company said that it’s fine with them, so long as they are identified clearly as ads. But they won’t be added to the network’s Ad Library, which allows anyone to track who is spending what on political advertising. Why not? Facebook wouldn’t say. Presumably because they are sort of ads, but sort of not ads. Also, these influencer ads can be fact-checked, unless they happen to contain the speech of a politician, in which case they can’t be fact-checked, because of Facebook’s existing policy of not fact-checking political ads. Does all that make sense to you? If so, then you might be in line for a job regulating political advertising at Facebook. If it doesn’t make sense, welcome to the rest of the human race.

Here’s more on the Bloomberg campaign and the platforms:

Manipulative: Influencer ads aren’t the only ways that Bloomberg is testing the limits of what the social networks allow. His campaign also posted a video clip of him asking during a recent debate whether any of the other candidates had started a business, followed by silence and quizzical looks from Joe Biden, Pete Buttigieg and several other candidates. The clip was edited, however — there was no long silence after Bloomberg asked the question, and the quizzical looks were taken from elsewhere in the program. Twitter said the video would be labeled as “manipulated” under its new policy, but Facebook said the edited clip was just fine.

Destroying norms: The New York Times wrote that the Bloomberg campaign is “testing the boundaries of what platforms like Twitter and Facebook allow in politics [and] they’re having trouble coming up with an answer.” Emerson Brooking, a resident fellow at the Atlantic Council’s Digital Forensic Research Lab, which studies disinformation, said that the campaign is “destroying norms that we will never get back,” and that it has “revealed the vulnerabilities that still exist in our social media platforms even after major reforms.”

Polluting: In an opinion column for The Guardian, Julia Carrie Wong writes that the Bloomberg campaign is “polluting the Internet” with its manipulative social-media tactics. “Mike Bloomberg’s campaign has spent the last month unapologetically performing the digital equivalent of dumping buckets of fresh garbage into the trash fire that is internet discourse in 2020, apparently with little or no concern for the toxic side effects,” Wong says. “It’s a cynical approach, and if left unchecked, it threatens to poison the atmosphere for good.”

Deepfakes: Although Facebook said it had no problem with the Bloomberg video that had been edited to make it look as though his opponents were speechless after his question, the social network has implemented new rules on “deepfakes,” or manipulated videos, after being widely criticized for not taking down a video of Speaker Nancy Pelosi that had been slowed down to make her appear drunk. The new policy says Facebook will not allow video that “has been edited in ways that are not apparent to an average person, and would likely mislead an average person to believe that a subject of the video said words that they did not say.”

Other notable stories:

Donald Trump has filed a lawsuit against the New York Times over an opinion piece the newspaper ran in the spring of 2019 by Max Frankel, who was the executive editor of the Times from 1986 to 1994. The column was entitled The Real Trump-Russia Quid Pro Quo, and argued that Trump had offered Russia a new foreign policy in return for help with the election. The lawsuit says the statements in the column were “false and defamatory,” and that the Times knew this when they were published. The newspaper said that the Trump campaign “has turned to the courts to try to punish an opinion writer for having an opinion they find unacceptable. Fortunately, the law protects the right of Americans to express their judgments and conclusions, especially about events of public importance.”

ABC News suspended one of its veteran correspondents, David Wright, after a video clip surfaced in which he made critical comments about Disney’s ownership of ABC and speculated that the network was not covering Donald Trump closely enough. The video was recorded surreptitiously by Project Veritas, a conservative group run by James O’Keefe that specializes in such ambush videos. Washington Post media blogger Eric Wemple said that by suspending its correspondent over such innocuous comments, ABC News “handed Project Veritas a win.”

Amanda Darrach writes for CJR about how a lot of the coverage of the new coronavirus, COVID-19, makes use of the same racist tropes about China and the Chinese that became prevalent in the wake of the SARS outbreak in 2002. Many of the stories about the new virus have focused on the allegedly bizarre eating habits of Chinese citizens (the virus has reportedly been traced back to a food market where wild animals are sold). A list in one story that included “unusual fare including live snakes, turtles, guinea pigs, badgers, and even wolf cubs” appears to have been designed to surprise or even horrify non-Chinese readers, Darrach writes.

The Tampa Bay Times said Wednesday that all full-time staffers will temporarily have their pay cut by 10 percent. The pay cut goes into effect next week and ends on June 5, according to a note distributed to staff and signed by the executive team. Five senior executives will take a 15 percent pay cut in that same time period, the note said. Management said the pay cut was “regrettable but necessary because revenues are falling short, a little in circulation and more seriously in advertising.”

Federal officials on Wednesday arrested several alleged members of a violent white supremacist group called Atomwaffen Division, including its two leaders, accusing them of plotting to intimidate journalists by calling police to their homes and offices and dropping off threatening fliers. John Cameron Denton, of Montgomery, Tex., is charged in Alexandria federal court with conspiring to call in fake threats targeting a ProPublica reporter and his office. Police arrived in force at both locations, at one point briefly detaining the reporter.

During the presidential debate on Tuesday night, a number of popular accounts were streaming and commenting on the broadcast on Twitch, Amazon’s real-time streaming service. Later that night, several of the channels — including Chapo Trap House and Mychal “Trihex” Jefferson — were suspended due to a copyright claim. After an investigation by the streaming service, however, all the channels were reinstated and Twitch apologized, saying the copyright claims were false. The service said none of the channels would be penalized due to the false claims.

Former Gawker writer Emily Gould has written a retrospective of her time with the defunct blog network in the mid-2000s. “It was part of my job to make fun of anyone who seemed to aspire to public recognition on any scale,” she writes. “Years before most people began to routinely broadcast their own personal details via social media, we told their stories for them, often in the least flattering way possible. I, of course, wanted to be recognized, too — recognized for the skillful, perceptive, and clever way that I made fun of these people.”

Former president Barack Obama on Wednesday called on South Carolina television stations to stop running an ad from a super PAC supporting President Trump that uses Obama’s words out of context in a misleading attack on former vice president Joe Biden. The Committee to Defend the President, a pro-Trump group, circulated an ad that falsely suggests that words Obama spoke in the narration of his own 1995 book were meant to describe Biden.

A former employee of The Roanoke Times has filed a federal lawsuit, accusing the newspaper’s parent companies of allowing sexual harassment to take place within the organization, discriminating against her on the basis of her sex and retaliating against her for bringing accusations of sexual harassment to light. Attorneys for Candace Lucas filed a 12-page complaint in U.S. District Court February 17. The suit names Times owners BH Media Group, Inc. and Lee Enterprises, Inc. as defendants and demands a jury trial to resolve the complaint.

Leave a Reply

Your email address will not be published. Required fields are marked *