Twitter 'Blacklists' Lead the Company Into Another Trump Supporter Conspiracy

On Wednesday, the greatest engine for bad takes on the internet screeched to a halt, briefly, as a far reaching hack and cryptocurrency scam hijacked many popular accounts, some with millions of followers like Bill Gates, Barack Obama, and Elon Musk.

In its effort to contain the damage, Twitter took the drastic step of temporarily disabling every verified Twitter account's ability to post, the logic being that if no verified account can post, not many people would be able to share the scam, even if they were hacked. For a moment, the blue checkmark crowd—journalists, celebrities, politicians, brands—were unable to tweet, and the unverified masses did little but hilariously celebrate their inability to tweet. It was a rare moment of true levity on the platform.

Then Twitter began fixing the problem, allowed everyone to tweet freely again, and the bad takes came roaring back. Most tiresome and toxic among them is the notion, held mostly by conservatives and Trump supporters, that Twitter was silencing them through the use of "blacklists." In reality, what we saw is Republicans stomping their feet at bare minimum, common sense moderation tools, a reaction that Twitter could probably mitigate in the future by being fully transparent about how its platform, and specifically its moderation tools, work.

This latest round of performed victimhood is based on screenshots Motherboard published in a story about the Twitter hack. According to our sources, and later confirmed by Twitter, hackers were able to take over accounts by "social engineering" of a Twitter employee, which gave them access to an internal tool. With this tool, the hackers were able to reset emails associated with the accounts and take them over.

The screenshots Motherboard published showed that, internally, some Twitter accounts are tagged as "Trends Blacklist" or "Search Blacklist." These are incredibly charged terms among conservatives, mostly because of a misleading 2018 tweet from Donald Trump which wrongly claimed Twitter was "shadow banning" prominent Republicans.

As we wrote after Trump's tweet, since the early days of the web, a “shadow ban” has been a moderation technique used to ban people from forums or message boards without alerting them that they’ve been banned. Typically, this means that a user can continue posting as normal, but their posts will be hidden from the rest of the community. The Twitter blacklists shown in screenshots we published do not prevent a user's followers or even the general public from seeing their tweets, but accounts put on these blacklists will be prevented from showing up on Twitter's "trending" page and will prevent them from showing up in search results.

Twitter explained this moderation tactic in 2018, after Trump's tweet. Twitter said it made changes to how it ranks its search results for accounts from what it considers to be "bad-faith actors who intend to manipulate or divide the conversation." A Twitter spokesperson told Motherboard on the phone that these blacklists are the same ones it explained in 2018 (though it didn't use the term "blacklist" at the time and has not used that word publicly.)

"We have very clear rules around trends and what we don't allow to trend," a Twitter spokesperson told Motherboard when asked about the "Blacklist" tags we see in screenshots of the internal moderation tool. Twitter also directed us to its "Twitter trends FAQs" page, where it makes clear the platform prevents content from trending if it contains profanity or adult/graphic reference, incites hate, or otherwise violates Twitter's rules.

Reached by phone, a Twitter spokesperson said the blacklist tags are "not new."

"We do disclose in this FAQ that accounts that violate the rules are prevented from trending," the spokesperson said. "This isn’t new and it’s not something that has been hidden, but it’s in the help center."

The Blacklist tags we see here are intended to prevent accounts that share hateful or graphic content, for example, from trending or appearing in search.

While this will be and already has become a major conservative talking point, we know for a fact conservatives are not being "silenced" or "shadow banned" because VICE's traffic tools show tweets by conservatives linking to our article, saying the screenshots we published reveal a shadow ban conspiracy, are leading thousands of viewers to our site.

People who are buying into a Twitter conspiracy are not completely wrong to be worried, however. While Twitter has explained before that these blacklists exist and the types of accounts that might get put on one, it does not alert users if they've been put onto a blacklist. It has not used the term "blacklist" publicly in this context, a term more loaded than "Twitter filters search results for quality tweets and results." ("Twitter may automatically remove accounts engaging in [rule breaking] behaviors from search" is a bit closer.) It has not explained if the blacklists are automated, administered by human moderators, or a mix of both, and it has not been specific about how or why an account it put on a "Trends Blacklist" or "Search Blacklist."

Earlier this month, a Twitter engineer led a charge within the company to make the code Twitter uses more inclusive; "blacklist" will become "denylist," and Twitter will stop using "Master/Slave" terminology as well.

We can't say with 100 percent certainty how the "Blacklist" tags work because we don't have full visibility into Twitter's moderation mechanism. We can see its public facing policy, but not debates inside the company, and more critically, the technical process by which accounts are suspended, banned, or prevented from appearing in search.

This opacity creates a situation where a devastating lapse in Twitter's security leads to a leaked image of an internal panel that contains the word "Blacklist," which Twitter doesn't use when talking about moderation, and which sounds more sinister than it is. For years, academics, journalists, and, yes, conservatives, have been demanding more transparency; these are the kinds of scandals that happen when a company's internal language is different from its carefully crafted blog posts and announcements.