Following the violence in Charlottesville, Internet businesses have been disassociating themselves from far-right political groups. PayPal decided to prohibit users from accepting donations “to promote hate, violence and intolerance”; 34 organizations were affected by the ban. Earlier, both GoDaddy and Google refused to host The Daily Stormer, a neo-Nazi site. Spotify announced that it would remove any streaming music “that favors hatred or incites violence against race, religion, sexuality or the like.” Dating site OkCupid deleted the profile of a white supremacist who was featured in a Vice TV segment about the Charlottesville, Va., rally, barring him for life.
Most Americans won’t worry much about neo-Nazis losing access to the public forum we call the Internet. Their ideas are repugnant, and we did fight a war against their kind not that long ago. End of story?
Not really. Experts, such as Cloudflare CEO Matthew Prince, say that for technical reasons a small number of tech companies may soon “largely determine what can and cannot be online.” Some argue that Silicon Valley is abusing its monopoly power over the Internet to suppress free speech. These critics would have Congress regulate the companies as a natural monopoly in order to restore basic rights.
Is there truly such a monopoly over the Internet? After reading that Facebook employees were suppressing conservative articles, Andrew Torba created Gab, a social-networking service open to voices from the right. Gab isn’t as popular as Facebook, but its existence does show that the dominant companies do not have a stranglehold on speech.
In any case, these decisions to deny service do not violate the freedom of speech. The First Amendment applies only to government actions. The man posting at 
Still, Americans in general and conservatives in particular have reason to suspect the tech companies might not be neutral toward content on the Internet. The James Damore case — in which a Google engineer circulated a memo suggesting the tech giant hired and promoted based on gender and race — indicates the leaders and employees of Google have strong leftwing political views. Indeed, Silicon Valley is known to lean to the left. It’s not implausible to imagine that censorship of certain views might ensue.
What then can be done to protect free speech without the courts?
If [internet companies] use their power to exclude in an arbitrary and political way, the nation will be worse off and the companies may suffer — and not just at the bottom line.
Markets will do part of the job. These companies are unlikely to deny service to mainstream political voices. After all, a person evicted from a service is no longer a paying customer, and their eviction might convince others to depart. Driving diversity critics out of Google would forego considerable revenue. Many people, not just extremists, have reasonable doubts about aspects of diversity policies.
Fear of the government will also constrain Internet censorship. Critics want to make Internet companies into public utilities because alt-right extremists have been denied service. Imagine what would happen if more legitimate voices were evicted from Google and the federal government responded by forcing the companies to behave better.
No one should want that. Law professor Danielle Citron has suggested several ways Internet companies can protect speech online.
First, affirm a commitment to American norms about free speech. The Internet giants operate globally, subjecting them to European regulations that offer less protection for free speech than does the United States. Europeans punish speech offensive to groups or religions (so-called “hate speech”) that we protect. American law draws the liberty line for speech at incitement to violence.
The companies mentioned promoting hate and violence as justification for evicting the alt-right. They seem to be following European norms. So far as the companies have discretion, they should affirm their support for the more liberal American norms. They should clearly define speech that will lead to being banned from a platform. That definition should focus on a close connection between problematic speech and violence and not on ambiguous terms like “hate speech.”
Second, companies should enact private due process for their regulation of speech. Clear definitions of the rules are essential in avoiding arbitrary and personal decisions about banning speech. The process of applying such rules should also be public. That could mean a formal public statement by the company of the rationale for banning some users of their social networks. It might also mean being public about how a company goes about identifying and prohibiting problematic speech. Such transparency about rules and methods will be open to public comment and inevitably criticism. It may also build trust among critics who fear arbitrary and politicized attacks on political speech.
Third, Internet companies should appoint an ombudsman to inform and report on their regulation of speech. They would act as a voice for free speech inside a company, a voice that should be dedicated to American norms on speech.
Internet companies are not the government. They can exclude speech from their domains without violating the First Amendment. But if they use their power to exclude in an arbitrary and political way, the nation will be worse off and the companies may suffer — and not just at the bottom line.
John Samples is vice president and publisher at the Cato Institute.