When you interact with someone on Twitter, is it really a person?
Does that ditzy celebrity really have that much influence?
Are those people on the internet even real?
These questions are increasingly at the heart of how people perceive Twitter. The social network boasts 330 million user accounts, which tweet about everything from social issues to the latest tech news, and represent everyone from Katy Perry to Pope Francis.
But a New York Times investigation found that millions of accounts on the service may be fake, created to help celebrities hawk products and to make "influencers" appear to have more -- well, influence.
"It's kind of like inflating your resume. They were buying status so they can have influence," said Jennifer Grygiel, a Syracuse University social media professor. And as if it wasn't obvious: "That's not OK."
It's not just inauthentic and slimy. The thought that millions of fake accounts are following, tweeting, retweeting and liking all sorts of stuff underscores the nagging existential question about what exactly Twitter is.
Over the past few years, Twitter's real-time, talk-to-anyone nature has helped give rise to social movements like #BlackLivesMatter and #MeToo. But it's also enabled a new type of hate campaign to flourish. Thousands of often-anonymous accounts swarm around topics or people they don't like, such as comedian Leslie Jones, outspoken video game developers or social activists like Renee Bracey Sherman.
What's increasingly become clear is that bots -- automated computer programs sometimes posing as real people online -- are at the crux of these issues. The New York Times report alleged that a company called Devumi sold millions of fake likes, retweets and followers to high-profile Twitter users including actor John Leguizamo, football great Ray Lewis and even Martha Lane Fox, a Twitter board member. (Leguizamo declined to comment to the paper; a representative for Lewis denied making Devumi purchases; and Lane Fox told the Times a "rogue employee" bought the bogus followers.)
Many of these accounts also appear to be stolen identities of real people, the Times reported.
Devumi didn't respond to a request for comment, as Twitter has said it's taking action against the company. Meanwhile, attorneys general from New York and Florida plan to investigate Devumi's practices.
"The internet should be one of the greatest tools for democracy -- but it's increasingly being turned into an opaque, pay-to-play playground," New York Attorney General Eric Schneiderman tweeted Saturday. He added that "impersonation and deception" are illegal.
The revelations about Devumi will likely be discussed in Congress and by law enforcement agencies. Both are investigating the extent to which Twitter and other tech sites were taken advantage of by Russian-linked agents during the 2016 US presidential election.
The most immediate fallout has shown up on Twitter. The Times reports that million of followers are now vanishing from prominent accounts.
Meanwhile, people are asking why the Times appears better at identifying bots than Twitter does.
Twitter declined to comment.
Emilio Ferrara, a research professor at the University of Southern California, said he finds it hard to believe Twitter took so long to deal with the issue.
"It's very strange they wouldn't do anything sooner," Ferrara said. "This is much deeper than just pointing out fake followers."
Ferrara also said there could be many more fake followers out there. A study he co-released last year with Indiana University found bots were behind as much as 15 percent of active Twitter accounts. Lawmakers have been using the study as a model. Twitter says about 5 percent of accounts are run by bots.
"It's hard to see the end to this right now," Ferrara said. "We're just at the tip of the iceberg."
iHate: CNET looks at how intolerance is taking over the internet.
Tech Enabled: How technology's playing a role in providing new kinds of accessibility.