In the three weeks since Mr. Musk cited fake accounts as a reason for putting his planned $44 billion acquisition of Twitter Inc. “on hold,” observers and participants in the deal have puzzled over the thinking behind the tycoon’s comments. The issue isn’t new for Mr. Musk, who has complained for years about Twitter’s ability to measure and manage automated accounts on the platform that often produce spam.
Whatever his intention in raising the issue, it is clear that Mr. Musk has had unusually extensive interactions with bots. As a habitual tweeter with more than 95 million followers, the Tesla Inc. CEO likely has far greater exposure and experience with fake and spam accounts than most on the social-media platform, researchers say. One estimate says spam, fake or inactive accounts make up the vast majority of his followers.
Mr. Musk is “an outlier among outliers,” said Darius Kazemi, a computer programmer who has spent a decade creating and studying bots and is currently a senior software engineer at Meedan, a technology nonprofit that aims to combat misinformation. “His experience is going to be different from not just the average user, but the average celebrity.”
Mr. Musk didn’t respond to requests for comment. Twitter declined to comment for this article but has defended its fight against spam accounts.
Researchers say Mr. Musk’s prominence and interaction with other users, as well as the topics he tweets about, make him a magnet for people looking to spread spam and other suspicious content. He has tweeted nearly every day for the past four years, and most of his tweets are mentions of or replies to other users, which encourages attention-seekers—human and bot alike—to reply to him.
Others with huge follower counts—such as former President Barack Obama, singer Justin Bieber and soccer player Cristiano Ronaldo—don’t tweet as often or engage as much.
Spam and fake accounts are an industrywide problem and can cause problems for advertisers and bad experiences for users. The accounts can be difficult to detect and are commonly managed by bots, which are computer programs that can automate posts and replies. Many bots have been programmed to achieve illicit goals, such as spreading false information and tricking people into spending money, tech and social-media analysts say, but others have more benign purposes, such as sharing news and weather alerts.
Mr. Musk’s sudden escalation of the bot issue last month raised suspicions among observers that he is using it as a negotiating tactic to lower the price amid the swooning market, or exit the takeover deal. That’s partly because he has complained about fake accounts on Twitter for years—long before he agreed to buy the platform in late April, as part of which he waived detailed due diligence on the deal. In 2018, he tweeted, “Lots of fake accounts on Twitter characterized by high following/follower ratio to make it seem like many real people when it isn’t. Wonder why.”
Mr. Musk’s feed is flooded with tweets hawking products, requesting money and pushing political agendas from what may be bots, coordinated groups or individuals. Replies to a recent tweet from Mr. Musk about computer programming, for example, touted new digital currencies for pet owners and the lesbian, gay, bisexual, transgender and queer community. Some bots are transparent, including one whose bio says it is automated to explain tweets in simpler words.
Outside estimates of the share of bogus accounts on Twitter vary—and all suffer from a lack of access to the company’s data. But whatever the overall percentage, industry experts say it’s most likely far higher for Mr. Musk.
Around 70% of Mr. Musk’s followers on Twitter are spam, fake or inactive, versus 41% for all other accounts with between 65 million and 120 million followers, according to an estimate last month from SparkToro LLC, a maker of audience-research software. Across the company’s data set, the average Twitter user has fewer than 100 followers and fewer than 10% are fake or spam accounts, the study showed.
Mr. Musk is “attracting so much attention that people may be trying to use him to find targets to scam,” said Kaicheng Yang, a computer scientist and Ph.D. candidate at Indiana University who researches bot activity on social media and helped build a bot-detection tool called Botometer. “If I’m a scammer, would I look at the average user? No.”
Twitter says it battles fake and spam accounts aggressively and suspends and blocks millions of accounts each week. It has long estimated that false or spam accounts represent fewer than 5% of its monetizable daily active usage or users—the number, most recently pegged at 229 million, that it emphasizes to advertisers.
Mr. Musk has questioned that estimate and voiced suspicion that the real share is far higher. He recently called Twitter’s rules “very bot friendly.”
Twitter has stood by its tally. “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),” Twitter’s chief executive, Parag Agrawal, tweeted last month in response to Mr. Musk’s remarks about putting the deal on hold due to doubts about the company’s bot estimates. Mr. Musk responded by tweeting a poop emoji.
“Anyone who uses Twitter is well aware that the comment threads are full of spam, scam and just a lot of fake accounts,” Mr. Musk said at a mid-May tech conference called the All-In Summit.
Marketing executives have said advertisers are aware of the presence of automated and spam accounts on social media—not just Twitter—and factor that into their strategies and measurements.
Whatever Mr. Musk’s intention, bad bots and fake accounts can pose a threat on Twitter such as by making an issue look like it is trending or influencing behavior, tech and social media experts say.
“Suddenly something that was fake popular becomes real popular,” said Sandy Carielli, a principal analyst at Forrester Research Inc. “While I would not speculate about Musk’s motivation for elevating the problem, he’s right it’s a problem.”
