Before Saturday, Oct. 27, relatively few people were familiar with Gab, the fringe social network developed as an alternative to Twitter and Facebook. By the end of the day, however, media outlets from Reuters to CNN had issued “explainers” on Gab as a online haven for far-right extremism online. The New York Times published a profile of the site by the following day. The surge of attention cast Gab’s future into doubt as PayPal pulled its service from the platform; then Gab lost its hosting provider, and next, its domain name service provider. As of Oct. 29, anyone visiting gab.ai is greeted with an announcement from CEO Andrew Torba stating that Gab has been forced offline: The site, he wrote, “has been no-platformed by essential internet infrastructure providers at every level. We are the most censored, smeared, and no-platformed startup in history.”
Gab’s users have posted their fair share of racist, anti-Semitic and all-around hateful content over the two years of Gab’s existence. But what led to Gab’s downfall was a single post by Robert Bowers, the man charged by the Justice Department for murdering 11 people in a Pittsburgh synagogue on the morning of Oct. 27. Minutes before his act of violence, Bowers, whose Gab account was full of viciously anti-Semitic posts, wrote on the platform: “I can’t sit by and watch my people get slaughtered. Screw your optics, I’m going in.”
For payment processing services PayPal and Stripe, along with Gab’s hosting provider—Joyent—and its domain-name provider—GoDaddy—this was enough to pull service. The decisions—what former Facebook chief security officer Alex Stamos described as “everybody stepping back during a trust fall”—are a window into the shift we’re living through of greater involvement by technology companies in policing the content that appears using their services.
We’ve seen this movie before. In August 2017, after the violent white supremacist rally in Charlottesville that resulted in the murder of a young woman, a string of domain-name services pulled hosting from the neo-Nazi website, the Daily Stormer. The website had published an article crowing over the death of the murdered woman, Heather Heyer: “Woman Killed in Road Rage Incident was a Fat, Childless 32-Year-Old Slut.” Days later, the content delivery network and security service provider Cloudflare—famous, as the tech site the Verge wrote, “for never banning anything”—kicked the Daily Stormer off its service as well, removing the protection that had allowed the site to weather the distributed denial of service attacks targeted against it as a form of vigilante justice. The Daily Stormer vanished from the web, though it quickly resurfaced on a series of ever-more-obscure top-level domains. (It currently exists as dailystormer.name after a long period of time on .red, courtesy of a group of service providers whose origins and motives remain difficult to untangle.)
The Daily Stormer’s exile prompted both furious debate and tortured introspection, especially within the technology community. Was it really right for a single company to have the ability to kick a website off the internet, no matter how terrible that website was? Sure, the people behind the Daily Stormer were neo-Nazis, but had GoDaddy and Cloudflare’s decisions started the internet down a slippery slope? Cloudflare CEO Matthew Prince wrote in an anguished email to staff (which soon became a public statement): “Literally, I woke up in a bad mood and decided someone shouldn’t be allowed on the Internet. No one should have that power.”
Gab does appear to have been a Cloudflare customer, and Cloudflare does not seem to have pulled its services from the platform along with GoDaddy and Joyent: Up until late on the evening of Sunday, Oct. 28, when Gab vanished from the internet, gab.ai was producing Cloudflare error messages. When I asked Cloudflare about the status of its relationship with Gab, the company declined to comment on Gab specifically but wrote that, “We've been vocal about that fact that deep infrastructure companies like Cloudflare should not be in the position to make editorial decisions based on content,” pointing me toward Prince’s 2017 statement on his decision to terminate service to the Daily Stormer. Cloudflare’s hand-wringing over its role in deciding what content is seen online is more unusual these days.
Gab presents itself as a social media platform analogous to Twitter, but it has a well-earned reputation as a cesspool for far-right extremism and harassment that makes its banning comparable to that of the Daily Stormer. (April Glaser, who reports on technology at Slate, shared a story of Gab’s CEO instigating harassment against her as she tried to report on the company.) Nevertheless, it’s striking how much calmer the reaction was to Gab’s demise within the community of technologists and tech-minded journalists than it was to the Daily Stormer’s exile. Writing to staff on Aug. 16, 2017, Prince noted the comment of one Cloudflare employee when informed of Prince’s plan to pull service to the neo-Nazi publication: “Is this the day the internet dies?” There was little of that this time around.
The phlegmatic response was not universal. “Totally happy to see racist platforms like Gab shrivel up and die,” computer scientist Matt Blaze wrote on Twitter, “but uneasy about payment processors being the arbiters and instruments of their death … nothing about this is simple or obvious to me.” But the tone has definitely shifted. Many fewer people seem much less concerned.
Blaze’s point is well taken. What many people—including myself—found troubling about Cloudflare’s banning of the Daily Stormer was not the idea of a platform moderating content at all, but where Cloudflare is located in the “stack” of infrastructure that keeps a website online. If a user is banned from a platform like Twitter for repeatedly posting violent anti-Semitic memes, there are plenty of other platforms—like, say, Gab—where that user can go. But if Cloudflare refuses to work with a website, there are relatively few comparable services that site can turn to. Cloudflare works at a more tectonic level of the internet. In the wake of Charlottesville, the alt-right has worked to effectively build its own internet from the ground up in order to solve this problem, but its success has been limited by the sheer magnitude of the task. Gab was, until recently, one of the project’s more successful outposts.
Cloudflare’s decision to ban the Daily Stormer from its service was a telling moment for an American cultural environment that had already become much less friendly toward technology, and much more willing to raise alarm over the potential dangers cultivated by the internet, in the wake of Russian electoral interference in the 2016 election. The relative lack of anguish over the downfall of Gab is similarly telling—a sign of just how far out from shore this tidal shift has taken us. I’m among those who were alarmed by Prince’s decision at the time, and I confess I was much less alarmed by GoDaddy and Joyent’s decisions. My colleague Matt Tait, no slouch when it comes to defending free speech online, expressed a similar change of mind.
So what changed? It’s true that Gab is far less well-known than the Daily Stormer, a website that had become notorious by the time it was forced offline. But the reaction when Gab went down was relatively muted even among the people who had long been familiar with the platform. More significant is the additional year’s worth of actual far-right violence and threats of violence in the United States since Heyer’s death in Charlottesville under a president who seems unwilling or incapable of putting his foot on the brake of his own incendiary rhetoric.
If Russian interference in the 2016 election moved generally tech-friendly groups within the United States toward skepticism and even fear over what the internet could do, the time between Charlottesville and Pittsburgh has underlined just how thin the line might be between speech threatening or implying violence and violence itself—and the role that platforms seem to play in providing, advertently or inadvertently, a space for violent or conspiratorial speech to fester into something much worse. In a sense, Gab’s refusal to moderate its own platform pushed the responsibility down the stack, to the point where Joyent and GoDaddy were the only entities in a position to cut off the firehose of harassment and anti-Semitism.
The immense pressure on Twitter to scale up its content moderation efforts and crack down on harassment and misinformation has moved it a long way since Tony Wang, Twitter’s U.K. general manager, characterized the platform as “the free speech wing of the free speech party” in 2012. That shift was one of the motivating reasons for Gab’s creation in the first place. But as of Sunday, Gab is no longer on the internet, and the tech community is mostly saying “good riddance.” And this change of heart is noteworthy in and of itself.