By tweeting, “Start your engines…,” late on Thursday, writer Matt Taibbi revealed the second leak of Twitter data.
Journalist Bari Weiss later tweeted about the documents in a thread headlined “The Twitter Files, Part Two: Twitter’s Hidden Blacklists,” which make public the existence of “secret blacklists” kept by Twitter workers.
As a result of how it handles certain information and accounts on its network, Twitter has come under fire recently. Now, the second release of the Twitter Files has confirmed what people have been saying for a long time: that the company has secret blacklists that groups of employees use to censor information from accounts they find offensive. Most of the censorship is political.
Twitter may restrict the display of certain accounts or popular topics by using these blacklists without notifying the individuals who would be impacted.
This information has led to more criticism of the company’s ways of doing business and the possibility that content control could be biased.
Elon Musk then told me that people who want to run for office or are already in office were also on the blacklist.
Blacklists were created and utilized by teams of Twitter workers to stop certain material from trending, according to Bari Weiss’ Twitter thread about the second release of the Twitter Files. Users have no means of knowing whether their material is being restricted and was not informed if their accounts were being censored in this manner.
Here’s an excerpt from the thread: “A new #TwitterFiles investigation exposes that teams of Twitter personnel establish blacklists, block disfavored tweets from trending, and deliberately restrict the visibility of individual accounts or even hot topics in private without alerting users.”
Twitter previously had the goal of “giving everyone the ability to develop and share ideas and information quickly, without limits,” but along the way, obstacles were still put in place, according to Weiss.
According to Weiss, during the pandemic lockdowns, Twitter blocked Dr. Jay Bhattacharya, a well-known Stanford professor, from posting on the social media platform. Bhattacharya opposed lockdowns because of their detrimental effects on children and was one of the first to criticize them.
In Bari Weiss’ article about the second release of the Twitter Files, it was shown that conservative pundits Charlie Kirk and Dan Bongino were also on the blacklist. This limited their reach and made it hard for people to read their tweets.
Weiss said that Twitter “denied that it performs such activities.” We do not shadow ban, Twitter officials Vijaya Gadde (then Head of Legal Policy and Trust) and Kayvon Beykpour (Head of Product) said in 2018. They also added: “And we definitely do not shadow ban based on political beliefs or ideology.”
Weiss said that what many people refer to as “shadow banning” is really what Twitter executives and staff refer to as “Visibility Filtering,” or “VF.” Several high-level sources corroborated this. “The term “VF” relates to Twitter’s management of user visibility. It employed VF to prevent searches for certain individuals, restrict how widely a tweet might be discovered, prevent some users’ postings from ever making the “trending” page, and prevent those posts from being included in hashtag searches.
She continued, “All without the users’ awareness.”
“We have a lot of influence over visibility. Additionally, we have a lot of influence over how your material is amplified. One Twitter developer informed us that “regular people don’t realize how much we do,” Weiss wrote. “Two more Twitter staff members confirmed.”
She went on to say that the moderating went “beyond the rank-and-file moderators implementing the company’s rules on paper,” disclosing the existence of the “Site Integrity Policy, Policy Escalation Support” (SIP-PES), which was made up of Twitter officials, including former legal director Vijaya Gadde, Trust & Safety chief Yoel Roth, and CEOs Jack Dorsey and Parag Agrawal, “and others.”
Weiss said that Twitter staff she contacted stated that the accounts affected would be those with large follower numbers, for whom “there would be no ticket or anything.” “This is where the biggest, most politically sensitive judgments were taken,” Weiss added.
Popular TikTok user Libs of TikTok (LTT), who was added to the “Trends Blacklist” and given the warning “Do Not Take Action on User Without Consulting With SIP-PES,” was one of the high-profile accounts that were subject to that extra degree of scrutiny.
Six times in 2022, for a total of eight suspensions, Chaya Raichik, the account’s owner, was prevented from using the service for up to a week at a time.
Weiss noted that Libs of TikTok did not technically break the rule but claimed that Twitter repeatedly notified Raichik that she had been banned for doing something that violated the company’s policy against “hateful behavior.”
She said, “However, the committee noted that LTT has not actively participated in conduct in violation of the hateful conduct guideline in an internal SIP-PES report dated October 2022, after her sixth ban.”
Weiss stated, “The committee claimed her writings promoted online harassment of “hospitals and medical practitioners” by implying that “gender-affirming healthcare is similar to child abuse or grooming.”
Weiss went on to talk about how the suspension of the Libs of TikTok account was different from real violations of the platform’s terms of service, like when Raichik was doxed.
According to Weiss, who claimed that the image is still up, “A photo of her house with her address was released in a tweet that has gotten more than 10,000 likes.”
Before it was taken down by editors, Raichik’s private home address was revealed by Washington Post writer Taylor Lorenz, according to Rebel News.
Weiss says that a screenshot of Yoel Roth talking about how to use the platform’s spam-enforcement features to silence users shows that “Twitter employees talked about using technicalities to limit the display of tweets and topics” in internal Slack chats.
In the future, Roth tasked the company’s Health, Misinformation, Privacy, and Identity Research Team with providing evidence in favor of enhancing the site’s visible filtering options, which he termed “non-removal policy interventions,” and which include de-amplification and deactivating interactions.
When asked about the rationale behind many of the measures taken, Roth said, “If exposure to, say, misinformation actually causes damage, we should apply remediations that lessen exposure, and restricting the spread or virality of material is a smart way to achieve that.”
For other policy domains, he said, “We got Jack on board with deploying this for civic integrity in the short term, but we’re going to need to make a stronger argument to get this into our repertoire of policy remediations.”
After the second batch of the Twitter Files came out, Elon Musk said that political candidates were pushed out of the way.