Twitter insists it has bots handled, claims it blocks 1 million spammers every day

Illustration by Michele Doying / The Verge

Reporters were told on Thursday that over 1 million accounts were kicked off of its platform a day. That number is double the one CEO Parag Agrawal stated in a May thread. In recent months, the number of fake accounts and how they are dealt with by the company has become a point of contention, as Musk threatened to back away from his purchase of the company.

The situation was discussed with the benefit of data, facts and context. He said in the thread that the platform suspends over half a million accounts a day and locks up millions of accounts a week. He said that the 5 percent number was derived from multiple human reviews of thousands of accounts that were randomly reviewed. Private account data, such as the user's location, is used by the reviews.

We suspend over half a million spam accounts every day, usually before any of you even see them on Twitter. We also lock millions of accounts each week that we suspect may be spam – if they can’t pass human verification challenges (captchas, phone verification, etc).

— Parag Agrawal (@paraga) May 16, 2022

It wasn't immediately clear if Thursday's metric is an update to these numbers or if it's counting something else. According to The Guardian, users that aren't allowed to make accounts are included in the figure.

In April, the two companies struck a deal for Musk to take the company private. Musk claimed in May that the deal was on hold and that the user base could be 20 percent bots or more. The company was accused of not complying with its obligations under the merger agreement after he asked the company to prove that the 5 percent number it had reported was accurate.

It's hard to meaningfully do calculations on the number of bots outside of the company.

They have responded to Musk's accusations before. In June of this year, the company said it would give him access to the "firehose" platform that would allow him to run his own analysis of all the posts on the platform. The company announced that it was rolling out an overhauled version of its reporting system, which would allow users to flag accounts that may belong to the bad guys.