Date: February 17, 2025
Updated on August 18, 2025 to better reflect my current positions, among
other reasons. In the interest of transparency, the original is published
here.
If Bluesky let all the Nazis in to "strengthen the marketplace of ideas" and then a Nazi doxxed another off-site over a rock, would that be fucked up, or what?
Moderation services are usually vague about their actions to protect the privacy of those affected, and to prevent giving too much information to would-be ne'er-do-wells that break the rules. Nevertheless, this hasn't stopped many from complaining..
For the sake of brevity, I'm not going to cover any events that happened prior to when I joined, like the N word handle incident in 2023.
In late November 2024, many people who had interacted with a specific account had their own account suspended. This was apparently related to matters of child safety.
After Bluesky banned the offending user, they banned users who had recently interacted with said user; later, the official Bluesky Safety account would allude that the chain ban was a mistake.
I'm still not sure exactly what was taken down, but if it was explicit art featuring underage characters, it would track with their previous enforcement, though the very heavy handed actions taken during this incident make me wonder if it was real CSAM, or perhaps something that ran afoul of 18 U.S. Code § 1466A (in both scenarios, the offending material must be reported and those who circulated it are in breach of that law).
(Not to be confused with the atproto RSVP website)
Not too long after the previous banwave, a certain provocateur and alleged yellow journalist named Jesse Singal would join the site. He would remain mostly undetected until around the 7th of December, where he would be banned, unbanned, had the "Intolerance" label placed on his account, banned again, then unbanned once more. The label was removed but I don't remember if it was before or after he got banned the second time.
This was a big deal when it was going on, not least because of how it pushed people's Twitter trauma buttons. A few people actually left Bluesky because of this. Others declared they were going to block people who pay for the (still) upcoming Bluesky subscription in the same way they did for Twitter Blue.
Eventually, the official Bluesky Safety would respond in a thread with a lot of blunders, to say the least.
The implication of the official account's posts is that Singal was never supposed to have been banned; at least one of the bans was the imposter detection firing, presumably the label was because of "inexperienced" (I would personally call them "good") moderators.
During the Singal Saga™, two major events happened:
Two bots created to track the actions of Aaron Rodericks, the head of Bluesky Trust & Safety.
The first bot was called the Head of T&S Activity Tracker. It was banned sometime later and the creator made a statement after Aaron rather dismissively acknowledged the bot's existence. The second bot, and the one most people know about was called What's our head of T&S doing?. It got banned when Aaron liked a spam bot's post, the bot quote posted it and the quote post went viral. They also banned the bot's creator, though they did reinstate her account a few hours later.
The incident of the second bot and its creator getting banned threw more gasoline on the fire and kept the Singal smoke going for a while longer. For what it's worth, the CEO of Bluesky Social, PBC, Jay Graber, later personally responded to someone asking about the ban, essentially implying that it was more inexperienced moderators responsible. Juni to this day still demands a response from Bluesky for a reason why, even though the subtext of Jay's post was clear.
From a moderation standpoint, both bots were in violation of Bluesky's policies. Bots that track a user like the Aaron bots did is against rule 1a of the Bluesky Developer Guidelines as reposting posts that Aaron liked or posted generate bulk interactions that will give notifications to those users. It also violates rule 2a of the regular Community Guidelines given how the bots' sole purpose was to harass Aaron.
Eventually someone found an exploit in the Bluesky API where having a super long (>255 characters in violation of spec) display name would cause the Bluesky API to fail when it encounters such a user. This was then used to suppress the replies on Jesse Singal's posts until the user was banned. The exploit was fixed shortly after.
Suppressing Singal's posts like that is great and all, but had it not been fixed quickly, a bot farm using that exploit could have effectively brought down Bluesky.
For the record, I still believe Bluesky's decision to not ban Jesse
Singal is a bad decision, and even though things mostly worked out
with him getting blocked by literally everyone and him essentially giving
up, Bluesky should not solely rely on community self-moderation as a
strategy. Its reputation as a "lefist echo chamber" is keeping most of the
really terrible people out, but eventually super popular bigoted people
(think like J.K. Rowling, not Richard Spencer) will try to join, and if
the policy is to allow said individuals like pre-Elon Twitter did, it will
continue to generate Singal-style controversies and make people
move away to other platforms further agitate the userbase.
As of now, there isn't a major place people can go to escape endorsed bigotry outside of Bluesky and maybe Mastodon (it being a major place is debatable), which is a harrowing thought when you think about it. That means Bluesky can join them without much repercussion, because there are no alternatives. But does Bluesky really want to do that?
There is the possibility they may be compelled or even forced to by the FCC (or by legislation, remember the TikTok ban?), but they would first need to get the approval from the Supreme Court to override the First Amendment like that. That's not impossible under the current US administration, but all we can do is wait and see. If they do weaken the First Amendment like that though...
...that is my personal rubicon.
