Elon Musk’s Twitter Files Are Bait

He doesn’t actually want transparency on the social-media platform.

Elon Musk's face distorted on a computer screen
Erik Carter / The Atlantic; Getty

Last night, Elon Musk celebrated the release of a new entry in the “Twitter Files” series, which aims to … Well, that’s complicated. It’s a supposed transparency project from Musk that, to date, has included giving two independent writers access to internal Twitter communications, as well as to the company’s Slack channels. So far, they’ve produced two threads, each totaling about 30 tweets, purporting to show how Twitter’s executives have schemed and colluded to censor political speech for partisan gain. The tweets are breathless, alluding at various points to “chilling passages” and “secret groups” of executive decision makers.

This latest tranche, released with commentary by Bari Weiss, gave me a look at something I’ve always wanted to see: a portion, at least, of the social platform’s internal moderation systems. The screenshots show how users can be flagged in various ways by the site’s employees: We can see that some users are marked as high profile, for example, and that individuals have a strike count, which seems to reference how many times the account has been caught in an infraction of Twitter’s rules. Then there are the juicier labels, like Do Not Amplify and the ominous-sounding Trends Blacklist.

As a set of documents, the screenshots are fascinating. But the “Twitter Files” entries are sloppy, anecdotal, devoid of context, and, well, old news. (Neither Weiss nor Twitter immediately responded to requests for comment.) Weiss’s thread purports to show how Twitter restricted the reach of several large accounts, including those of the Stanford doctor Jay Bhattacharya, the right-wing activists Dan Bongino and Charlie Kirk, and Chaya Raichik, who operates the infamous Libs of TikTok account. Via the internal screenshots, it argues that Twitter, contrary to its public claims, shadowbanned conservative accounts through a process the company calls “visibility filtering.” Similarly, the thread implies that a committee of top Twitter executives known as “Site Integrity Policy, Policy Escalation Support” acts as a kind of content-moderation black site—an off-the-books cabal that makes off-the-books decisions on high-profile accounts. (Meta has recently been embroiled in a controversy over a similar “cross-check” system.)

The subjectivity of moderation decisions across the social web poses tremendous and complicated problems—which is precisely why journalists and academics have paid close attention to it for more than a decade. But there’s no secret conspiracy: These claims largely comport with what Twitter has publicly stated about its moderation policies over the past few years. Since May 2018, Twitter has noted that it will change how a user appears in the site’s search feature if that user behaves in a way that “detracts from healthy public conversation.” Twitter has an entire Frequently Asked Questions page dedicated to this type of de-amplification, which it says it does based on a combination of algorithms and human decision making, but not via political ideology.

I do not believe that Elon Musk cares about the thorny particulars of content moderation. By releasing these internal documents selectively, Musk gets to be on the offensive, whipping up a crowd that’s eager to cry censorship and shadowbanning. He’s not even interested in paying attention to the details: As the first mega-thread dropped, last weekend, Musk admitted that he hadn’t actually looked at the files yet—although that didn’t stop him from making boneheaded statements, like suggesting that its contents proved a large-scale First Amendment violation.

In fairness, the Twitter Files do show that the company makes amplification decisions about certain accounts. And while that’s not especially revelatory to people who’ve paid attention, the files do speak to the immense power wielded by tech platforms. It’s a power that makes Republicans and Democrats queasy, albeit often in different ways. As content-moderation experts will tell you, it’s a messy system where people can make the wrong call with occasionally disastrous results. In this way, the Twitter Files do what technology critics have long done: point out a mostly intractable problem that is at the heart of our societal decision to outsource broad swaths of our political discourse and news consumption to corporate platforms whose infrastructure and design were made for viral advertising.

The dynamics of the entire affair are Trumpian in the most exhausting way possible. Attempts to engage with the attention hijacker only further his purposes, but ignoring him can feel as if you are letting him get away with distortions and the last word. What Musk seems to really want is to anger liberals, delight his right-wing and reactionary bedfellows, and alienate the mainstream media. This is surely why he releases documents only to writers who’ve expressed alignment on his pet issues. It would also explain why he tweets disparagingly about former employees, accusing them of being politically biased just a month after urging his nearly 115 million followers at the time to vote for Republicans in the midterm elections. And it is why Musk castigates Twitter’s former leadership for its de-amplification policies while promoting a policy of “freedom of speech, but not freedom of reach,” which amounts to the exact same thing.

The trolling is paramount. When former Facebook CSO and Stanford Internet Observatory leader Alex Stamos asked whether Musk would consider implementing his detailed plan for “a trustworthy, neutral platform for political conversations around the world,” Musk responded, “You operate a propaganda platform.” Musk doesn’t appear to want to substantively engage on policy issues: He wants to be aggrieved.

Still, it’s possible that a shred of good could come from this ordeal. Musk says Twitter is working on a feature that will allow users to see if they’ve been de-amplified, and appeal. If it comes to pass, perhaps such an initiative could give users a better understanding of their place in the moderation process. Great! But Musk hasn’t given any reason to suspect that he cares about transparency, beyond wielding it as a weapon against his perceived enemies. He doesn’t engage in good faith with his critics, and he’s perfectly fine with a selective release of internal information that’s geared toward pleasing his fans. At times it seems his entire Twitter project is little more than identifying perceived injustices from site’s past—biases and so-called abuses of power—and using them as a playbook to run on his ideological opponents.

And so, what matters more than the content of the Twitter Files is the constant buildup, and the framing of their release as a bombshell act of unprecedented transparency. It signals indisputable proof that the bloodthirsty audience is about to see something big. As is so often the case online, the signal alone seems to be enough. No surprise, then, that the writers behind the Twitter Files and Musk himself are constantly teasing the next installment. The conspiracy that’s just out of reach is always more powerful than the one that was (supposedly) revealed.

Charlie Warzel is a staff writer at The Atlantic and the author of its newsletter Galaxy Brain, about technology, media, and big ideas. He can be reached via email.