No, this is the result of the laziness of social networks to look for automated solutions to solve at-scale (read: at the lowest cost) problems of their own making.

More generally this is the result of giving free rein on global discourse to unregulated companies, who ultimately took full advantage and profiteed off of it.

There's another way for Facebook and other social media companies to make sure this doesn't happen without employing stupid algorithms that censor people. Drop the bots, hire a bunch of people to pretend to be ISIS or Stormers, and ban[0] everyone they come into contact with. Leave everyone else alone.

They won't do this. Why?

The way that social media is structured - and the way Facebook et all make their money - is to get as many people as possible on platform, get them addicted to the platform, and then show ads. One of the easiest ways to do this is to provoke and generate outrage - as you can see on Twitter, which basically exists to turn people into public figures and then into "villains of the day". This is also why their systems reject context; because stripping speech of its context is the easiest way to construct a villain of the day.

The reality is that these outrage groups actually tend to be really, really small and close-knit. There's a small handful of people who actually feed the algorithm new extremist content, and everyone else parrots them without thinking much of it because they're angry. Extremism only looks large and prevalent because social media is designed to create echo chambers and manufacture consent. And extremists just so happen to be Facebook's best customers - people who already have outrage to play to, who will spend hours on platform, and so on. Of course they aren't going to ban their whales!

However, an ineffective bot that just randomly bans things that sound vaguely extremist-like? That gives you the appearance of Doing Something, without actually doing it, and it falls in line with the usual Silicon Valley protocol of "if it's not worth doing at scale, it's not worth doing at all". Treating extremism like a weirdly-shaped spam problem gives social media companies cover and is how we get stupid bots that think a Holocaust movie is Nazi propaganda.

[0] I do not consider taking down the content of jihadis or neo-nazis to be censorship. Jihadis and neo-nazis are groups with explicit, stated goals to do violence to groups of people for what they say or believe. An ISIS beheading video is not an artistic statement or a political diatribe, it is a threat to other Muslims. "Get in line or you'll be next." Allowing this to be spread around as if it were speech accomplishes the goal of censoring non-extremists.

If you want to argue that tankies or ANTIFA do this too, fine, but the poster I was replying to was specifically pointing out right-wing extremism.