To hear President Trump describe it, social media companies are hopelessly biased against him, his supporters and the political right.
He complained about that alleged bias in an interview with Fox News’s Tucker Carlson earlier this month.
“I know for a fact, I mean, a lot of people try and follow me, and it’s very hard,” Trump said about Twitter, where his account is offered as a recommended follow for new users. “I have so many people coming up that they say, ‘Sir, it’s so hard. They make it hard to follow.’ What they’re doing is wrong and possibly illegal. And a lot of things are being looked at right now.”
Trump told Carlson that “Facebook was against me. They were all against me. Twitter was against me."
This has been a slowly building line of rhetoric by Trump for about a year. Last July, conservatives were mad at Twitter for engaging in what critics described as “shadow-banning” people — not kicking them off the service but instead making their accounts harder to find. It became a rallying cry, with conservative Twitter users applying red X’s to their displayed usernames to complain about the practice.
Twitter was in fact changing how some accounts were displayed, but there’s no evidence it had anything to do with politics. Instead, following criticism of rampant abuse and racism on the platform, the company implemented a system that would downplay users who had been a frequent focus of complaints about abusive activity. In May 2018, the company “started using behavioral signals and machine learning to reduce people’s ability to detract from healthy public conversation on Twitter,” a company representative tweeted. The effect was that some accounts were removed from the mix and became harder to find — with the effect, Twitter hoped, of detoxifying conversations.
Similar patterns happened on other social networks. The 2016 election cycle marked the advent of Trump as a politician but also of social-media platforms as channels of significant influence. That included Russian interference efforts and the emergence of social-media-centric facets of the alt-right. It included fervent and often vitriolic political battles by supporters of the various presidential candidates.
As the election unfolded and after it was over — though not obviously because of the election — the networks decided they needed to clean things up. Google began to focus on ensuring it didn’t propagate fake news and worked to root out potential bias in its algorithms. The company has removed financial incentives for or shut down YouTube channels promoting false or inflammatory content. Facebook removed its trending news section and began advocating for moving more conversations into private groups.
As with Twitter, these efforts have been criticized by conservatives as having targeted them for political reasons. Google's efforts, for example, were the focus of a story by James O'Keefe's Project Veritas, which presented them as an effort to block Trump's reelection.
Among the conservatives making claims about bias is another Trump: Donald Trump Jr. After an Instagram story he posted was removed — it compared the need for a wall on the border with walls at a zoo — he became a champion of identifying anecdotal examples of social-media bias. In March, he wrote an opinion piece for the Hill cobbling together various incidents and alluding broadly to having heard of many other examples of conservatives being targeted.
His father’s administration, recognizing Trump Jr.'s ability to read the room, picked up this tactic. Earlier this year, it put out a call for anecdotal examples of people (read: conservatives) being targeted by social media companies. Such anecdotes don’t prove systemic problems, but that’s never been a barrier for Trump or his administration. For example, anecdotes are central to Trump’s rhetoric against immigrants even when data shows his arguments are wrong.
The White House’s request for anecdotes was the precursor for this week’s social media summit at the White House. It’s an event theoretically predicated on discussing the “opportunities and challenges” of the Web, as The Post reported on Tuesday.
Twitter, Facebook and Google didn’t receive invites. Instead, the White House invited a number of friendly social-media users and provocateurs to “engage directly” with those who’ve been the focus of bias on social-media networks. The invitations included ones to O’Keefe from Project Veritas and Twitter user @CarpeDonktum, whose pro-Trump videos and animations have been repeatedly shared by the president.
They also included cartoonist Ben Garrison, whose cartoons of a buff Trump conquering his various enemies are popular with the president. Garrison tweeted a picture of his invite.
In short order, some of Garrison’s past cartoons resurfaced. One in which he promoted the QAnon conspiracy theory, for example, or one in which he bought into conspiracy theories surrounding the shooting death of a Democratic National Committee staffer in 2016.
A cartoon in which Garrison drew national security staffers controlled by both George Soros and a family that’s long been a centerpiece of anti-Semitic conspiracy theories quickly made the cartoonist’s invitation a focus of criticism.
On Wednesday morning, an update from the White House: Garrison was no longer invited to the social media summit. Which, of course, is exactly the sort of decision-making by social-media companies that the summit was created to rail against.
Garrison appears not to have been one of the users who Twitter de-emphasized in its searches (the so-called “shadow banning"). He tweets his cartoons regularly, including the one that got him booted from the social media summit. It was fine for Twitter, but not for the White House.
It makes some sense that more Trump supporters might get caught in filters set up to catch abusive or derogatory content. After all, this is a president whose very first response to a debate question as a candidate was to defend his disparagement of women by claiming that he wouldn’t be “politically correct.” He and his supporters have continually argued that derogatory comments or jokes might only be unwelcome because of hyperactive political correctness; it’s not hard to see how that sensibility might tip over into more objectively abusive behavior.
An article in Vice News earlier this year interviewed a Twitter employee who claimed that the company was struggling with eradicating white nationalist content in part because developing filters that would remove that content might also sweep up rhetoric used by far-right politicians. How does an algorithm differentiate between Trump calling migrants from Mexico rapists in his campaign launch speech and a white nationalist making the same claim?
Trump’s summit is meant to argue that his supporters are being targeted systematically for their political beliefs. To make that argument, he’ll likely point to anecdotal examples that he claims proves his point. But he’ll be doing so a day after having done exactly what he claims to be objecting to: removing a voice for sharing offensive content.
If an anecdote proves systemic bias, as Trump seems to believe, we must come to an unwelcome conclusion: The Trump administration wants to silence pro-Trump cartoonists.