Twitter revealed that it has suspended hundreds of thousands of accounts in an effort “to combat violent extremism.”
So why does the service still feel like it’s chock full of extremist assholes?
The numbers, released on Tuesday in Twitter’s 10th semiannual Transparency Report, are striking.
Between Aug. 1, 2015, and Dec. 31, 2016, Twitter suspended 636,248 accounts with 376,890 of those suspensions coming in the second half of 2016 alone. Of the latter group, the report notes that internal teams were responsible for 74 percent of those suspensions, as opposed to users reporting content.
Overall, the number of extremism-related suspensions is on the rise. As Reuters points out, in the first half of 2016, there were 63,000 suspensions a month. In the same period the year before, there were 24,000.
Twitter reportedly has approximately 313 million monthly active users, meaning the banned accounts represent only a tiny fraction of the total user base. Still, though, it shows that management can be proactive in addressing terms of service violations.
And yet. White supremacists like Richard Spencer are still free to use the service (he was briefly suspended last year but later had his account reinstated), and countless other non-marquee creeps still harass women, minorities, and LGBTQ users on a daily basis.
In 2016, for example, Salesforce allegedly walked away from a rumored acquisition at least in part because of all the trolls.
A bit of sunshine does manage to peek through the dumpster-fire smoke, however. That Twitter was able to catch the aforementioned 74 percent of accounts it deemed extremist with “internal, proprietary spam-fighting tools” is a good sign. Basically, the company has shown it can handle a huge number of TOS violations on its own and that not every response to bile-spewing nonsense will require coordination with the FBI.
Indeed, the report actually documents several incidents of pushback against the government. Twitter notes it “received 88 court orders and other legal requests […] directing us to remove content posted by verified journalists and news outlet accounts.” The company said it didn’t do anything with the “great majority of these requests” excluding the ones originating in Turkey and Germany.
So there’s something at least.
Whether Twitter can use the lessons it’s presumably learned to make its service a slightly less flaming pile of Richard Spencer-esque nonsense remains to be seen. Here’s hoping.