Elon Musk may never be able to kick bots off Twitter

This article was first featured in Yahoo Finance Tech, a weekly newsletter highlighting our original content on the industry. Get it sent directly to your inbox every Wednesday by 4 p.m. ET. Subscribe

Wednesday, May 18, 2022

Getting rid of bots isn't as easy as it sounds

Elon Musk’s $44 billion deal to buy Twitter (TWTR) is in purgatory. The brash Tesla (TSLA) CEO put the purchase on hold May 13, saying, on Twitter, of course, he couldn’t move forward with the acquisition until he can verify that bots make up less than 5% or less of the social network’s total user accounts.

The billionaire, and richest person on Earth, has made taking down automated spam and scam bots, not to mention loosening speech moderation, the centerpiece of his buyout campaign. Musk claims that as many as 20% of Twitter’s users could be bots. According to Israeli tech firm Cyabra, bots could make up some 13.7% of Twitter's 229 million total accounts.

But it might be impossible to eliminate bots on Twitter, regardless of who’s in charge.

“I actually don't think it's possible,” explained Kai-Cheng Yang, a researcher at Indiana University’s Observatory on Social Media.

Bots, simple programs that automate tasks, are largely used to send out quick tweets. These tweets can alert you about a big football game or even let you know the time. But malicious actors can also use bots to scam people out of cryptocurrency or spread disinformation. Countries like Russia have been accused of using bots to spread false information about vaccines and U.S. elections. And unfortunately, these malicious bots have ever-evolving ways of evading detection.

FILE - Elon Musk speaks at the SATELLITE Conference and Exhibition March 9, 2020, in Washington. Musk's great Twitter turnabout, in which he disclosed his massive stake in the social media company, got a seat on its board, publicly floated drastic changes and then turned down the board role, all happened in a week. (AP Photo/Susan Walsh, File)
Tesla CEO Elon Musk has put his deal to purchase Twitter on hold. (AP Photo/Susan Walsh, File)

“It's an arms race between…us and the bot creators. We might come up with some solutions to identify such types of accounts and then they will try to use evolved techniques…to evade our detection. And then we come up with…some new methods to catch them and then they change their strategy again. We've been seeing this a lot,” Yang said.

In other words, Musk may not be able to get his wish of a bot-free Twitter even if he decides to eventually move forward with the deal again.

Twitter makes building bots easy, but they’re not all bad

Musk has been tangling with Twitter bots since at least 2018. Scammers have previously deployed bots masquerading as the CEO in an attempt to trick users into turning over cryptocurrency or handing over their crypto wallets.

And, notoriously, this past January a Twitter bot created by a then-19-year-old named Jack Sweeney tracked the activity of Musk’s private jet. Musk said the bot posed a security risk, and offered to pay Sweeney $5,000 to take it down. Sweeney refused.

Musk has been accused of exploiting bot situation to get out of buying Twitter. After all, Twitter has reported that bots make up less than 5% of its total users every quarter since at least 2014. In other words, Musk should have known about Twitter’s bot estimates before making his bid for the company.

Not all bots are bad bots, either. Even Sweeney’s bot simply used publicly available information to track Musk’s flights. In fact, Twitter purposely makes it easy to create bots using an API, or application programming interface. Using an API, users can create bots that tell you the weather or how your favorite baseball team is doing.

“Some bots are designed specifically for malicious purposes, but in general bots are neither good nor bad,” explained Ari Lightman, professor of digital media and marketing at Carnegie Mellon University’s Heinz College.

“Twitter for a long period of time has had things like amplification bots, messaging bots, magnitude bots,” Lightman said. “These are all bots designed to either accelerate a message, amplify a message, send out a message at a routine time, those sorts of things.”

The problem is that malicious actors often figure out ways to use bots to spread disinformation, or scam and harass other Twitter users.

“It's like a constant battle,” Lightman said. “Research gets published associated with how to detect malicious intent associated with some bots, and then folks with malicious intent utilize that information to see how they can…avoid detection and scanning.”

Bots will exist even if Twitter tries to end them

But what if Twitter decided to just stop offering APIs for bots? Wouldn’t that go a long way toward ending the malicious bot problem? Not necessarily. See, Twitter’s API makes it easy for users to create bots. However, scammers have other ways to create bots.

They can hijack existing user accounts or use a web browser to automate bot behaviors. It certainly takes more effort than a standard API-based bot, but experienced cybercriminals and disinformation specialists can definitely pull it off, creating phony accounts they can easily control.

“As long as there's a financial incentive, there will be bots,” said NYU Tandon School of Engineering associate professor Damon McCoy.

“As long as someone can make money off of the bots, the bots will exist. End of story. If someone can, in a capitalist society, make money off of something and it's not horribly illegal, they're going to do it.”

Need proof that bots, like the dinosaurs in “Jurassic Park,” can thrive without APIs? Look no further than Facebook (FB) or Instagram or TikTok, each of which regularly contend with bots trying to spread disinformation.

There’s also the possibility that Musk could force Twitter users to prove they’re human by verifying who they are. But that would take away the anonymity of the platform and imperil political dissidents who live in countries with oppressive regimes.

Even adding a series of steps to the log-in process proving users aren't a robot, would be a tough sell, according to Yang.

The reason? Well, people are generally lazy. And adding extra friction points to the sign-up or log-in process could turn off a not insignificant number of users.

In the end, even if Musk buys Twitter, which is increasingly in doubt, it’s unlikely he’ll ever be able to kill off bots entirely.

By Daniel Howley, tech editor at Yahoo Finance. Follow him @DanielHowley

Read the latest financial and business news from Yahoo Finance

Follow Yahoo Finance on Twitter, Instagram, YouTube, Facebook, Flipboard, and LinkedIn