Twitter’s latest anti-abuse tools will keep eggs out of your mentions

Twitter’s long, halting, and frustrating road to reducing abuse has arrived somewhere useful. The company said today it would introduce new optional filters that will prevent you from being notified when certain accounts tweet at you. This includes accounts without profile photos — called “eggs” for the default egg picture given to new Twitter accounts — as well as accounts without verified phone numbers or email addresses.

It may sound like a minor thing, but the move shows Twitter moving gradually to increase the friction for anyone who wants to use the platform for spam and abuse. By forcing users to upload photos and verify phone numbers and email addresses if they hope to be seen by their targets, Twitter is making would-be harassers do a little more work when they create new troll accounts. And making harassers do more work is one of the best tools we have to reduce harassment. “Our platform supports the freedom to share any viewpoint,” the company said in a blog post, “but if an account continues to repeatedly violate the Twitter Rules, we will consider taking further action.”

Twitter is augmenting its new filters with machine learning. When Twitter detects that accounts are repeatedly tweeting at accounts that it does not follow, or engaging in unspecified other activities likely to signal abuse, Twitter will temporarily limit their access to the platform. (During this time a suspended account’s tweets are visible only to the people that follow it; Twitter won’t specify all the instances in which it will suspend an account for fear that trolls will successfully game the system.)

The nice thing about this second change is that it begins removing the onus of reporting harassment from the victim. Twitter trolls have a broad but finite range of negative behaviors, and beginning to chip away with them using machine learning represents one of the most user-friendly changes Twitter has made in years.

In another user-friendly step, the company says it will try to provide more information for people who do report harassment on the platform. One of the more frustrating things about Twitter’s abuse reporting process is that it so rarely results in an action being taken; Twitter says it will attempt to communicate more. People who file reports will be notified when Twitter receives their report and when the company takes action against the reported account; those notifications will be visible in the notifications tab.

And one more: you can now mute keywords from the timeline, and best of all, you can mute them for specific amounts of time. This isn’t strictly an abuse thing — you’re seeing lots of tweets about the same person, you can now just put that person’s name on a 24-hour timeout and they’ll come back to your timeline when that time is over. I called on Twitter to introduced time-based muting three years ago, when it first unveiled a mute button. I wanted it for accounts and keywords, but this is a start.

After years of ineffective posturing about abuse — and a failed attempt at selling itself — Twitter announced a campaign to reduce harassment in January. So far, it has resulted in temporary shadow bans and more suspensions for users that attack verified accounts. It has also resulted in some missteps, as when Twitter was forced to roll back some anti-abuse tools after complaints that they encouraged abuse.

Still, if Twitter has done a better thing for its users in the past year, it’s news to me. The changes will roll out globally in coming days, the company said.

Correction, 9:01 p.m.: This article has been updated to reflect the fact that the changes only enables time-based muting for keywords, not accounts.