Someone has to sort out the bots from the humans on Twitter and we’ve got a team on it.

We launched a label for the good ones, and while the label isn't an endorsement, we want you to know which bots are made by developers to help, entertain, or just harmlessly share info. 🧵

What even constitutes a "good bot" is subjective (depending on what you find useful or funny), but the robot icon identifies automated accounts so there’s no mistaking it for a human. The automated label even links you to the creator of the bot.

Here's how that looks: https://t.co/NERyeydicC

@Tinycarebot & @queerlitbot are just a few of the good ones around here.

Unfortunately, not all of them are like that, and with the help of a forensic team of investigators, machine learning is helping us sniff out the bad bots.

When it comes to bad (spammy, deceitful) bots, we're focused on what IMPACT they have on conversations & working on mitigating that impact.

We're proactively sniffing them out & you can help by reporting suspicious accounts. Some things to consider: https://t.co/P9bXowi5Im

1- Don't judge an account by its name. Twitter autogenerates handles with first names & a string of numbers until a user changes that, which some people simply never choose to do.

2- Bots may Tweet a lot, but so do many people. We all use Twitter in different ways so we can't assume that an account is a bot because they love to Tweet.

3- Just because someone has crappy opinions, it doesn't mean they are a bot (even if it really feels like it). Poor taste? Maybe. But they are not necessarily a bot just because they have hot takes.

Thing is, the bots may be Tweeting away but not influencing real-life behavior. Either way, when the Tweets are abusive or violate rules, the Platform Integrity team works to remove them ASAP.

The good bot label will help with identifying cool accounts that may enrich your experience on Twitter. Of course, it is totally up to you if you follow them or not.

Read more here: https://t.co/KKRi8Co27T