`No fakery please': Twitter suspends over 5M spam accounts every day

Twitter suspends over half a million spam accounts every day, usually before anyone can see them.

"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)," said Twitter CEO Parag Agarwal.

On Monday, Parag took to Twitter to explain how the company tackles spam accounts.

He explained that spam accounts are not just binary (human/non-human). In a lot of cases, the most advanced spam campaigns use combinations of coordinated humans and automation. In a nutshell, real accounts are compromised and used to advance spam campaigns. This makes them sophisticated and hard to catch.

"The hard challenge is that many accounts which look fake superficially – are real people. And some of the spam accounts which are the most dangerous – and cause the most harm to our users – can look legitimate on the surface," Parag explained.

Spam accounts are always evolving to tackle various software. "Thus, it is possible that a set of rules built to detect spam today might not do the same tomorrow," Parag said.

While Twitter is working towards removing as many spam accounts as possible every day, many spam accounts slip through.

Parag said that the spam accounts, in the last four quarters, are less than 5 % of Twitter's Monetisable Daily Active Users (mDAUs)

Each human review is based on Twitter rules that define spam and platform manipulation and uses both public and private data (eg, IP address, phone number, geolocation, client/browser signatures, what the account does when it's active…) to decide on each account.

"The use of private data is particularly important to avoid misclassifying users who are real. FirstnameBunchOfNumbers with no profile pic and odd tweets might seem like a bot or spam to you, but behind the scenes, we often see multiple indicators that it's a real person," Parag said.

"There are LOTS of details that are very important underneath this high-level description. We shared an overview of the estimation process with Elon a week ago and look forward to continuing the conversation with him, and all of you," he said.

Parag explained the above in response to Elon Musk's comments on the spam accounts and the Twitter algorithms. Elon Musk, who has agreed to buy Twitter for $44 billion, said he was putting the deal "on hold" until he had more information about how the company measures bot accounts.

Three days ago, Elon tweeted that he will do a random sample of 100 followers on Twitter. Parag contradicted Elon's statements by stating that 'multiple human reviews (in replicate) of thousands of accounts, that are sampled at random, consistently over time.

"Unfortunately, we don't believe that this specific estimation can be performed externally, given the critical need to use both public and private information (which we can't share). Externally, it's not even possible to know which accounts are counted as mDAUs on any given day," Parag said.