Artist and comedian Leon Chang, known online as leyawn, was handed two "stikes" by BlueSky over a post that violated anti-discrimination guidelines protecting groups with "protected characteristics" such as "age, asylum-seeking/refugee status, caste, disability, disease ethnicity, gender identity, immigrant status, religious affiliation, sex, or sexual orientation."
Leon's post, though, was not about any of these characteristics. It was about federal law enforcement:
imagine showing up to your office job and there's a crowd of people in the lobby heckling you every morning. "no one likes your emails" "your tie is ugly" lol. anyways i think everyone who works for ice should kill themselves
Under federal law, cops—including secret ones wearing masks while executing Americans in the street—aren't a protected class. Some states include police under hate crime legislation concerning specific crimes, but that's not quite the same thing.
The antidiscrimination law in Washington state, where BlueSky is headquartered, does include military employment as a protected class. Even then, though, ICE is a branch of the Department of Homeland Security, not the U.S. military, and the law as written specifies housing, financial services, and public accomodation; social media is not any of those things (not even the latter.) Moreover, Bluesky hasn't added "wears camo in Minneapolis" to the list in a policy that plainly echoes the legal terminology of federal and state law. Working for ICE is a choice, not an immutable characteristic.
Moderation on social media is extremely difficult (cf. Masnick's Theorem), certainly at scale, especially when you're recreating Twitter from last principles.
moderation is, inherently, a subjective practice. Despite some people's desire to have content moderation be more scientific and objective, that's impossible. By definition, content moderation is always going to rely on judgment calls, and many of the judgment calls will end up in gray areas where lots of people's opinions may differ greatly. Indeed, one of the problems of content moderation that we've highlighted over the years is that to make good decisions you often need a tremendous amount of context, and there's simply no way to adequately provide that at scale in a manner that actually works.
Over time, though, a consistent direction of failure can become apparent. Perhaps they fear backlash from higher forces and a loss of control. Capricious moderation doesn't typically flow from conscious ideological bias at odds with policy (ideological bias tends to be openly expressed in policy) but because the environment encourages hypervigilant policing of content that approaches insitutional anxieties. In this case, presumably, the comical suggestion of violence ("should kill themselves") poked the Bluesky Moderation Team's political nightmares and the result was a false "strike" for discrimination.
