Maryland Smith’s David A. Kirsch, who is researching pro-Tesla Twitter bots, or “fanbots,” says irony pervades Elon Musk’s stated reasoning to abandon his $44 billion offer to buy Twitter. While the company has filed suit to enforce the deal, Musk says his decision is based on the platform’s traffic of bots, or fake accounts.
“There’s something deeply ironic about the person who may have profited more than anybody from the existence of bots on Twitter using the existence of bots on Twitter as an excuse to back out. The hypocrisy in that argument should be lost on no one,” says Kirsch, associate professor of strategy and entrepreneurship at the University of Maryland’s Robert H. Smith School of Business.
In a working paper (summarized here) with co-author Mohsen Chowdhury, Kirsch identifies a set of non-human accounts called fanbots and explores the possibility that these accounts may have influenced the trajectory of Twitter discourse about Tesla, the principal source of Musk’s wealth.
Musk’s decision to abandon his bid to acquire Twitter follows his disagreement with the company over how much data the social network should share to satisfy Musk’s inquiries about the bots, which are created by software to automatically scan the social media platform for specific activity, then act to amplify or discredit certain messages or news.
“Musk’s general argument – that somehow he was surprised by the extent of bots on Twitter – doesn’t pass the laugh test because even if Musk didn’t order the fanbots that we observed, he’s known they were there – in many cases for seven or more years – or longer than we’ve had computational corporate propaganda,” Kirsch says. “We date the public debate about computational propaganda to its political manifestations in the 2016 elections here and in the U.K., but we found corporate activity predated political propaganda by several years.”
As for Twitter’s limited disclosure of bot traffic, Kirsch says the company has, to some degree, benefited from the existence of bots, and it’s not really in their interest to reduce the bot-inflated number of monthly active users. “If they really wanted to get rid of every bot, I think they could do that quite easily,” he says. “It’s not that hard. But it would open them up to even more accusations of censorship because there would surely be some autonomous-seeming human users that their bot-detection algorithm would identify as bots.”
