from the reality-bites dept
For all the talk of the “Twitter Files,” as we’ve detailed, they’ve mostly been, at best, misleading, and frequently actively wrong. One of the big reveals, we were told, was that the Files were going to expose the political machinations of how Twitter banned former President Trump. And, indeed, Bari Weiss’s “Part Five” of the Twitter Files, back in mid-December, purported to reveal the big secret reckoning. But if you haven’t heard much about it since then, it’s because… they were a complete flop when it came to anything of interest. Basically, it was exactly what some of us said the day it happened: a difficult decision with a number of competing factors going into it. One that could have gone either way, but recognizing the gravity of what happened on January 6th, and the genuine concern that Trump would continue to whip his fans into an insurrectionist frenzy, one that you can see a reasonable argument for making.
And while Musk (falsely) insisted that the big reveal was that Trump didn’t actually violate Twitter’s policies, that’s also a misreading of what happened. What we’ve learned is that Trump and other Republican leaders were actually given special treatment over the years, because they tended to violate policies way more often than Democrats. But, knowing that Republicans would flop to the ground and fake injury any time they were faced with even having to take the slightest bit of responsibility for violating policies, all the big social media platforms went above and beyond to better protect the high profile accounts of Republican rule breakers.
And while many people tried to paint the decision to finally ban Trump as some sort of “proof” that the company leadership was a bunch of left-leaning censors, the reality seemed to be quite different. Even Weiss’ big reveal was simply that there was strong and heated internal debate about what to do, with many employees (mostly not directly engaged in content moderation issues) calling for the company to ban him, while executives and trust & safety folks questioning whether or not that would be appropriate.
Right at the end of last year, though, as the House Select Committee investigated January 6th was wrapping up, some of the details of what they discovered about Twitter’s debate was leaked to Rolling Stone, and presents an even more detailed picture of how the company strongly resisted calls to ban Trump.
In the draft summary, written by the committee’s “purple” or social media team, staffers were more pointed about what they saw as the failures of big social media companies.
“The sheer scale of Republican post-election rage paralyzed decisionmakers at Twitter and Facebook, who feared political reprisals if they took strong action,” the summary concluded.
The report shows that, again contrary to the public narrative pushed by Musk and friends, Twitter’s leadership wasn’t as deeply engaged in the various political happenings:
And even days after the insurrection, former Twitter employees told the committee that executives were still slow to recognize the risk Trump could pose in inciting future violence. After Trump tweeted that he would not attend Joe Biden’s inauguration, Safety Team employees testified that they saw “the exact same rhetoric and the exact same language that had led up to January 6th popping underneath” his tweets, leading to fears of another act of mass violence.
Some of the people who worked on that social media report, separately wrote an article for Tech Policy Press, talking about some of what they saw, which didn’t make it into any public report. They note that their research debunked the widely held notion that the social media companies acted with their bottom line in mind in refusing to limit disinformation, and again found that fear of angering Republicans was a key motivating factor:
At the outset of the investigation, we believed we might find evidence that large platforms like Facebook, Twitter, and YouTube resisted taking proactive steps to limit the spread of violent and misleading content during the election out of concern for their profit margins. These large platforms ultimately derive revenue from keeping users engaged with their respective services so that they can show those users more advertisements. Analysts have argued that this business model rewards and incentivizes divisive, negative, misleading, and sometimes hateful or violent content. It would make sense, then, that platforms had reason to pull punches out of concern for their bottom line.
While it is possible this is true more generally, our investigation found little direct evidence for this motivation in the context of the 2020 election. Advocates for bold action within these companies – such as Facebook’s “break glass” measures or Twitter’s policies for handling implicit incitement to violence – were more likely to meet resistance for political reasons than explicitly financial ones.
As the report’s researchers found, Twitter was extremely resistant to putting in place policies that might make Republicans mad:
For example, after President Trump told the Proud Boys to “stand back and stand by”’ during the first presidential debate in 2020, implicit and explicit calls for violence spread across Twitter. Former members of Twitter’s Trust and Safety team told the Select Committee that a draft policy to address such coded language was blocked by then-Vice President for Trust & Safety Del Harvey because she believed some of the more implicit phrases, like “locked and loaded,” could refer to self-defense. The phrase was much discussed in internal policy debates, but it was not chosen out of thin air – it was frequently invoked following the shooting by Kyle Rittenhouse in Kenosha the previous summer. But the fact it appeared in only a small fraction of the hundreds of tweets used to inform the policy led staff to the conclusion that Harvey’s decision was meant to avoid a controversial crackdown on violent speech among right-wing users. Ironically, elements of this policy were later used to guide the removal of a crescendo of violent tweets during the January 6th attack when the Trust & Safety team was forced to act without leadership from their manager, whose directive to them was, according to one witness, to “stop the insurrection.”
The authors noted, explicitly, that people reading the Twitter Files to say that Twitter was controlled by a bunch of coastal liberals trying to silence conservatives have it quite backwards:
One clear conclusion from our investigation is that proponents of the recently released “Twitter Files,” who claim that platform suspensions of the former President are evidence of anti-conservative bias, have it completely backward. Platforms did not hold Trump to a higher standard by removing his account after January 6th. Rather, for years they wrote rules to avoid holding him and his supporters accountable; it took an attempted coup d’état for them to change course. Evidence and testimony provided by members of Twitter’s Trust & Safety team make clear that those arguing Trump was held to an unfair double standard are willfully neglecting or overlooking the significance of January 6th in the context of his ban from major platforms. In the words of one Twitter employee who came forward to the Committee, if Trump had been “any other user on Twitter, he would have been permanently suspended a very long time ago.”
None of this should be a surprise to anyone who has been reading Techdirt throughout all of this. For years, we’ve pointed out that the whining from “conservatives” that social media was biased against them was nothing more than an attempt to “work the refs” and basically lean on the decision makers to make sure the opposite was true. It was designed to make sure that the trust & safety teams at these companies were so frightened about the potential for politicians and the media to make a big deal out of any decision that it effectively gave them free rein to ignore the rules and push the boundaries, and the companies (beyond just Twitter) were too scared of the potential reaction to react.
This is especially ironic, given all the nonsense we’re hearing now about how the FBI was supposedly “censoring” people via Twitter. The truth is that it was actually Republican politicians, media, and influencers who scared Twitter away from taking actions against rule violators who were deemed to be prominent conservatives.
Filed Under: content moderation, donald trump, january 6th, january 6th committee, working the refs
Companies: twitter
