Elon Musk wants to know that spam bots are actually less than 5% of Twitter users

Since 2013, Twitter has reduced the proliferation of fake accounts on its platform, holding that “fake or spam” accounts account for less than 5% of its user base, although independent researchers say the number could be three times higher.

The discrepancy could now affect the outcome of Elon Musk’s $ 44 billion cash deal when the billionaire tweeted on Friday that the takeover bid was “temporarily on hold” when he asked for information on the number of fake Twitter accounts.

While this may be a little more than a discussion strategy of Mask, according to current employees and independent social media researchers, almost nothing is certain about how these accounts are defined or distributed.

The social media platform said in a May 2 public filing that less than 5% of its 229 million daily active users who are targeted with ads are “false or spam” based on an internal review of their accounts. It did not specify how the image was responsible for the automated, parody and pseudonymous profiles allowed on the platform.

Twitter did not respond to a request for comment.

Researchers estimate that 9% to 15% of millions of Twitter profiles from 2017 are based on a recent study by a firm that monitors automated accounts or bots and online conversations based on a preliminary study.

“They have underestimated that number,” said Dan Brahmi, CEO of the Israeli technology company Cyber, which uses machine learning to detect fake accounts.

Cybra estimates that the percentage of unverified Twitter profiles is 13.7%.

Questions have been circulating on all social media platforms about the role of bots in spreading misinformation since 2016 when Russia interfered in the US presidential election to boost Donald Trump’s candidacy and harm his opponent Hillary Clinton.

Meta, which owns the competing platforms Facebook and Instagram, similarly estimates that fake accounts represent about 5% of monthly active users on Facebook, according to the latest data for the fourth quarter of 2021. Meta further estimates that about 11% of “duplicate” accounts in which a single user maintains multiple accounts, a practice considered acceptable on Twitter.

Twitter’s rules prohibit camouflage and spam, which means “fake” accounts are banned if the company determines that their purpose is to “deceive or manipulate others”, for example, engaging in scams, coordinating abuse campaigns or artificially increasing engagement.

Spam clearing out

Over the years, Twitter has invested in clearing spam accounts. In 2018, Twitter acquired a company called Smyte, which specializes in spam prevention, security and safety. Twitter has removed “spammy and suspicious accounts” in an effort to improve the platform’s health, reducing its user base to 1 million in July 2018 and reducing its stock.

Filippo Menzer, a researcher at the Indiana University Observatory on social media, says Twitter has become more aggressive in removing such unverified accounts, although the nature of the threat is evolving and difficult to measure.

With an integrated network controlled by both humans and software and so-called cyborg accounts, “manipulation has become more sophisticated,” the manager said, adding that these bad actors could “flood the network and then delete their content to avoid identification.”

Even though the numbers are actually small, according to researchers, bots can have a big impact and a big impact in the form of a handful of online conversations.

A Carnegie Mellon University study analyzing the spread of COVID-19 lies in 2020 found that 82% of the top 50 influential retweeters were bots.

Inside Twitter, measuring and identifying false or spam accounts is a complex issue that is not well defined or understood by the company’s own employees, three sources familiar with the matter told Reuters on condition of anonymity.

Twitter uses a variety of metrics and definitions to measure such accounts, which also depends on the company’s accuracy in identifying spam-producing content, one of the sources said. It also challenges the ability to accurately estimate the number of fake and spam accounts – and new accounts are always being created, the source said.