The journalist claims that Twitter has created secret blacklists to limit the visibility of certain tweets and some right-wing accounts.

The *Twitter Files* is a collection of excerpts from internal Twitter communications posted on the social network by journalists including Weiss and Matt Taibbi. The name Twitter Files (*Twitter files*) was created by Elon Musk himself. Taibbi does not specify how he obtained the documents, but there is little doubt that they were passed on to him by Musk, who promised to go public with internal discussions on Twitter on a number of topics.

The first part of The Twitter Files focused on how the social network blocked the distribution of an article from the conservative New York Post tabloid, in the middle of the American election campaign, in October 2020, which focused on the documents from the laptop. Owned by Hunter Biden, son of Joe Biden. At the time, the New York Post published excerpts of Hunter Biden’s emails suggesting he was using his father’s political connections to do business in Ukraine. Photos of a sexual nature or showing Hunter Biden using drugs were also posted online and presented as coming from the same computer.

At the time, Twitter founder Jack Dorsey was quick to dismiss the social network’s communication on the matter: Our communication around our actions in the @nypost article was not very good. And block URL sharing via tweet or DM without explaining why we’re blocked: it’s unacceptable, it’s forwarding explanations posted via one of bluebird’s official accounts. The images in the articles include personal and private information, such as email addresses and phone numbers, which violates our rules, the teams allege.

The second part highlights potentially more controversial practices.

Weiss posted screenshots of internal Twitter tools that moderators can use to limit the accessibility of posts and accounts. The account of Charlie Kirk, a young conservative activist from Turning Point USA (a frequent source of misinformation about Covid and elections) had a “not to amplify” sign.

Bongino, a right-wing talk show host, has been blacklisted, according to documents seen by Weiss. A Stanford professor who promoted the herd immunity program in the early days of Covid was also blacklisted from trends by Twitter moderators.

Weiss seems most concerned about the multiple suspensions of the Libs of Tik Tok account, which Twitter has indirectly (but not explicitly) sanctioned for violating Twitter’s *hateful conduct* policy. The account is popular for highlighting members of the LGBTQ community who have been sued and harassed by online trolls.

The conservative journalist also describes these content moderation practices (and the internal debates surrounding them) as “negative and secretive.” But the Twitter feed lacks vital context. More importantly, it doesn’t address Twitter’s policy announced in 2018 that it would begin restricting content it felt “stealth” the conversation.

Weiss also doesn’t explain how Twitter’s previous content moderation decisions conflict with the current policy announced for the company under Musk’s leadership, insisting that Twitter will uphold freedom of speech but not the freedom to enter into a commitment to curtail offensive content. Notably, Musk went so far as to suspend Kanye West again for posting hateful content.

Importantly, Weiss only highlights controversial moderation decisions involving right-wing figures. This echoes the ideological pattern presented by Taibbi (he only pointed to Twitter reactions to the Biden campaign, although he acknowledged that the Trump White House has also sought to remove controversial content that was respected by top Twitter executives).

The setup (the conservative CEO who leaked the documents to right-wing journalists) gives the impression that the #TwitterFiles project isn’t so much about getting to the bottom of serious wrongdoing as it is about stoking the passions of self-righteous right-wingers. fall prey to the bias of big tech companies.

Even Twitter founder Jack Dorsey argued that a more transparent approach that cut out the middleman would serve the public better. If the goal is transparency to build trust, why not post everything unfiltered and let people judge for themselves? , he asked Musk on Twitter.

The words of the journalist

Here are his comments on his Twitter page:

A new #TwitterFiles investigation shows that groups of Twitter employees create blacklists, block disliked tweets from trending, and actively limit the visibility of entire accounts or even trending topics.

Twitter was once on a mission to empower everyone to create and share ideas and information instantly without barriers. Despite this, obstacles have been erected along the road. Take Stanford’s Dr Jay Bhattacharya ( @DrJBhattacharya ), for example, who has argued that Covid lockdowns will harm children. Twitter secretly blacklisted him from trending, which prevented his tweets from trending.

Or consider Dan Bongino (@dbongino), the popular right-wing talk show host who was once blacklisted from the search engine.

Twitter has set the account of conservative activist Charlie Kirk (@charliekirk11) to “Do Not Boost”.

Twitter has done no such thing. In 2018, Twitter’s Vijaya Gadde (then head of rights and trust policy) and Kayvon Beikpour (head of product) said: We don’t use shadow bans. They added: And we are certainly not seeking a shadow ban based on political views or ideology.

What many people call a shadow ban, Twitter executives and staff call it a visibility filter, or VF. A number of high-level sources have confirmed its meaning. Think of a view filter as a way to hide what people see at different levels. It’s a very powerful tool,” a senior Twitter employee told us. VF refers to Twitter’s control over user visibility. Twitter used VF to block individual user searches; limit the scope of detection of a certain tweet*; to prevent certain users’ posts from appearing on the trending page*; and inclusion in hashtag searches. All this without the knowledge of the user.

We control the appearance somewhat. And we have some control over the amplification of your content. A Twitter engineer told us that normal people don’t know what we do. Confirmation from two other Twitter employees.

The group that decided to limit coverage to certain users was the Strategic Response Team – Global Escalation Team, or SRT-GET. He often handled up to 200 cases a day. But there was a level beyond the official administration, the moderators who followed company policy on paper. This is *Site Integrity Policy, Policy Escalation Support*, also known as *SIP-PES*. This secretive group included the head of legal, policy and trust (Vijaya Gadde), the global head of trust and security (Yoel Roth), the later CEOs Jack Dorsey and Parag Agrawal, and others.

The most important and politically sensitive decisions were made here. Think high follower counts, controversy, another Twitter employee told us. There would be no trace or anything else for them.”

One of the accounts that reached this level of scrutiny was @libsoftiktok, which was blacklisted from trending and designated as Do Not Act User without consulting SIP-PES.

Raichik says the account (Chaya Raichik launched in November 2020 and currently has more than 1.4 million followers) has had six suspensions in 2022 alone. each time Raichik was blocked for a week. Twitter has repeatedly said it suspended Raichik for violating Twitter’s policy against hateful conduct.

But in an internal SIP-PES memo in October 2022, after its seventh suspension, the committee acknowledged that LTT had not engaged in conduct that directly contravened the “hateful conduct” policy. See here*:

The committee justified its suspension internally, arguing that its posts encouraged online harassment of hospitals and medical providers, saying that gender-affirming health care was child abuse or care. Contrast this with what happened on November 21, 2022, when Raichik himself was doxxed. The tweet, which received more than 10,000 likes, included a photo of his home with his address.

When Raichik told Twitter that his address had been leaked, he said Twitter support responded with this message*: We reviewed the reported content and found no violations of Twitter’s rules. No action was taken. The doxxing tweet is still in place.

In internal Slack messages, Twitter employees talked about using methods to limit the visibility of tweets and threads. Here’s Yoel Roth, then global head of trust and security at Twitter, in a direct message to a colleague in early 2021:

Six days later, in a direct message with a staff member of the Health, Misinformation, Privacy and Identity research group, Roth called for more research to support the expansion of non-policy interventions. .

Roth wrote: The assumption underlying much of what we do is that if, for example, exposure to misinformation causes direct harm, we should employ remedial measures that reduce exposure, and limiting the spread/virality of content is a good way to do that. He added: “We’ve convinced Jack to do this for short-term civic integrity, but we’re going to have to make a stronger case for making this part of our repertoire of policy tools — particularly other policy areas.”

Source: Barry Weiss

And you?

how do you read
Do you share Jack Dorsey’s remark to Elon Musk: If the goal is transparency to build trust, why not publish everything unfiltered and let people judge for themselves?