Not All Bots Are Bad, and Twitter Knows It

They’ve been alleged to have helped shift the course of the 2016 presidential election , and are a key component of Elon Musk’s ongoing attempts to wriggle out of his $44 billion purchase of Twitter. They’re lambasted as the single thing ruining social media , and lauded as a key weapon in state-sponsored cyberwarfare. Bots have become public enemy number one in recent years, and remain in the spotlight thanks to Twitter’s former head of security Peiter “Mudge” Zadko’s claims that the company’s “senior management had no appetite to properly measure the prevalence of bot accounts.

” Twitter’s apparent inaction in tackling bots—and its supposed underestimation of the proportion of users that are bots—is one of Musk’s main arguments for attempting to sidestep his deal to buy the social media platform. But are bots all that bad? “The use of the term ‘bot’ causes a lot of confusion for folks,” says Christopher Bouzy of Bot Sentinel, which tracks inauthentic behavior on Twitter. “The media has done a disservice in that regard.

” By lumping together useful automated accounts that track political missteps—such as those that monitor when politicians delete tweets or alter Wikipedia pages about themselves—and state-sponsored, inauthentic accounts that exist solely to push a disruptive line about the world into the single definition of ‘bot,’ we’re doing a disservice to those automated accounts designed for good. “We try to tell folks in the media to start talking about ‘inauthentic accounts’ or ‘fake accounts,’” says Bouzy. “But don’t use the term ‘bots.

’” The US Department of Homeland Security’s Office of Cyber and Infrastructure Analysis sums up the duality of bots: They “can be used on social media platforms to do various useful and malicious tasks while simulating human behavior. ” Those useful tasks include ensuring transparency around world leaders and major businesspeople. Politwoops was one of the first examples of social media bots designed to maintain accountability in public office, set up in March 2011 by web developer Breyten Ernsting to track tweets deleted by politicians.

(Ernsting did not respond to a request for comment. ) A UK version focused on parliamentarians, Tweets MP Delete , arrived two years later, with localized versions in New Zealand , South Africa , Ireland , Belgium , Pakistan , and Germany . Bots also keep track of edits to Wikipedia pages by IP addresses linked to parliaments in countries including the UK , US , and Australia .

Just this week, Parliament WikiEdits, the UK version, spotted an attempted cleanup of the Wikipedia page of Kwasi Kwarteng, the UK chancellor, by someone with a UK parliament IP. The edits to Kwarteng’s Wikipedia page removed a section suggesting the politician had put undue pressure on a parliamentary standards watchdog investigating a colleague over “serious breaches” of lobbying rules. Bots are also generating news in other areas.

College student Jack Sweeney made a name for himself by running @ElonJet , which tracks the movements of Elon Musk’s private jet. He’s since gone on to create 30 different Twitter bots tracking the movements of jets, including those linked to Russian oligarchs and tech executives such as Jeff Bezos and Mark Zuckerberg. But bots don’t even need to be useful.

“I don’t think bots have to do something worthwhile,” says V Buckenham, founder of Cheap Bots, Done Quick! , a free tool that helps people create automated Twitter accounts. Tens of thousands of bots have been developed using the platform, most of which Buckenham says aren’t useful. “It’s a joyful thing or a creative thing,” they say.

“It’s a form of creative expression, whether that be something lots of people are following, or something that just amuses you. ” Some bots blur the boundary between utility and diversion. Journalist Karen K.

Ho began posting reminders for people to put down their phones and stop doomscrolling through Twitter at the start of the pandemic. “I had developed quite a following during the pandemic because—understandably—many people were doomscrolling for information on how to deal with the coronavirus pandemic,” she says. She was doing so manually, typing out the missives and hitting send, until she began to find it tiring to do so—particularly late at night, when people were most likely to aimlessly flick through Twitter.

So she built a bot to do the job for her. @doomscroll_bot now tweets every hour, reminding people to log off, alongside sitting better and not slouching. It’s followed by nearly 90,000 people.

“I think of bots as a type of medium, or a tool of the internet,” says Ho. And Ho believes that such innocent, useful bots aren’t necessarily conducive to success. “What I do with my bot doesn’t feed capitalism,” she says.

“With disinformation bots, people can make money. That’s why they exist. ” Part of the issue, says Buckenham, is that the term “bot” has an elastic meaning.

A 2021 academic paper shows that using three different methods of defining inauthentic behavior on Twitter results in three dramatically different estimates of the proportion of users. Buckenham says that people point to new Twitter users, who often have a string of numbers automatically assigned in their username, as being state-sponsored. “It’s a filter bubble thing,” says Buckenham.

“Different people use Twitter in wildly different ways. You may only see people who tweet in a similar way to you, so when you encounter people using the service in a different way, you assume they’re fake or illegitimate. ” What one person perceives as a Russian-sponsored bot designed to sew disinformation could in fact be a middle American mom who isn’t bothered about changing her username from the default option given to her when she signed up.

Buckenham believes the shift from bots being a neutral word to a loaded one occurred in 2016, when bots became the bogeyman that supposedly won Donald Trump the US presidential election . It signaled a change from the denomination of bots as something that corners of the internet like Weird Twitter would use, to a tool of disinformation designed to seed chaos and, in time, polarize society. Such polarization has continued through to Musk’s approach to Twitter bots, which have been presented as the enemy of a harmonious platform.

That’s not the case, says Buckenham. “They add serendipity and beauty to the timeline,” they say, pointing to bots like BoschBot , which dutifully posts small sections of Hieronymus Bosch paintings every few minutes. Buckenham created a similar bot of their own, @softlandscapes , which posts generated pastel-colored landscapes every six hours.

It’s one of their most popular bots. “Mainly it’s there because you follow it, and among all the doom and gloom and terrible stuff that happens on Twitter, you see a beautiful, calming landscape,” they say. “It takes you out and distracts you from all the stressful things in everyday life.

”.