Twitter started, as with so many other Silicon Valley success stories, as a side hustle.
In 2006, Jack Dorsey was working at the podcasting platform Odeo, when he got the idea for an SMS-based way of keeping tabs on friends via status updates. He started building out the system, sending the first tweet on March 21, 2006, and it has since turned into one of the most influential social networks in the world, with an estimated 330 million monthly active users (MAU) — or so it claims.
But with the release of a new investigative report from the New York Times on Saturday about the black market for fake social media followers, that number is, along with so much else that we thought we knew about the Internet, questionable.
Here are five things we learned from the New York Times report:
1. The number of fake accounts on Facebook and Twitter are higher — much higher — than you think.
By some counts, up to 48 million of Twitter’s claimed 330 million monthly active users are actually fake. Twitter, of course, has denied this, but the platform is far from alone.
Back in November, Facebook told investors that up to 60 million fake accounts were on the platform, more than twice as much as previously estimated. These fake accounts, also known as bots, are influential in shaping public opinion, amplifying messages, and spreading fake news.
Meanwhile, Devumi, one of the most popular companies for buying social media followers and the subject of the Times investigation, alone has 3.5 million fake accounts on offer for its clients.
2. Bots are everywhere, but basic design choices on Twitter make them easier to operate on the platform.
Unlike many other commercial sites, signing up for a Twitter account is very easy — for humans and computers alike. Accounts do not need to be associated with real, verifiable names — for better or worse.
Additionally, registering on Twitter doesn’t require an anti-spam check, which is one easy barrier that websites put up to deter bots. The Times article quotes a former Twitter engineer, Leslie Miley, who worked on security and user safety, as saying that, “Twitter as a social network was designed with almost no accountability.”
3. A lot (more than you think) of your favorite celebrities have bought followers.
The article lists a diverse group of celebrities and influencers that have purchased followers, including model-turned-entrepreneur Kathy Ireland, who has purchased over 750,000 followers (of a total 1 million + follower count); Michael Dell, the founder of Dell Computers; and Martha Lane Fox, a Twitter board member that purchased 25,000 followers right after joining the company’s board. When confronted, some of the celebrities have shifted the blame to assistants or PR agencies, others have refused to respond to comment, while still others apologized — and saying that everyone does it.
Dean Leal, who works in the adult film industry and tweets from @PornoDan, told The Times, “Countless public figures, companies, music acts, etc. purchase followers. If Twitter was to purge everyone who did so there would be hardly any of them on it.”
Almost all celebrities denied knowing they were buying fake accounts though. James Cracknell, a British rower and Olympic gold medalist, acknowledged that the whole system was “fraud.” He bought 50,000 followers on the service. “People who judge by how many likes or how many followers, it’s not a healthy thing.”
4. There is a whole supply chain in the market of social media bots.
The Times investigation found that Devumi was not creating fake followers from scratch, but rather, purchasing them from “wholesalers” that connect anonymous bot makers from around the world to the retailers like Devumi. These wholesale sites are often much less user-friendly — but also significantly cheaper. One of the sites offers 1,000 “high-quality, English language bots with photos” for less than a $1, while Devumi charges $17 for that same amount. The $16 difference is how businesses like Devumi — and the wider supply chain of fake accounts — have become so lucrative their money.
It was only in the aftermath of the elections that the platform started dealing with fake accounts, as it had previously been focused on spam and abuse from people.
5. Real social media identities are getting stolen.
The bot-makers aren’t just inventing people from scratch but, often, impersonating unsuspecting social media users, by changing one letter in their username, for example, — from i to l — which is almost impossible to spot at a glance. This has real consequences for the impersonated individuals, since the fake accounts will retweet whatever clients have “bought” them — including questionable fake news and adult film content.
One of the women impersonated, Salle Ingle, a 40-year-old engineer in Colorado, told the Times that she’s worried about the effect of the fake account on her employment prospects. “I’ve been applying for new jobs, and I’m really grateful that no one saw this account and thought it was me,” she said.