A Co-Founder of the Digital Rights Group EFF Has Died, and Every Big Problem He Started Worrying About 28 Years Ago Is Getting Worse

State surveillance, digital privacy, a closed-off internet… the problems he worried about have all come to pass.

John Perry Barlow, who died this past Thursday, lived a full life.

Because of his family's faith, Barlow was forbidden to watch television or read books and magazines until the sixth grade, kept away from pop culture. But at 15 he met Bob Weir, who would become the singer and guitarist of Grateful Dead, and he wrote countless songs for the band, which eventually entered the Rock and Roll Hall of Fame. Grateful Dead was active in the 1960s, and its songs carried the counterculture of the time.

Barlow studied at Harvard Law School, served as state campaign manager for Dick Cheney, later George W. Bush's vice president, ran a ranch to pay off family debts, wrote screenplays and made films, roomed with the ever-controversial Silicon Valley entrepreneur Sean Parker, and even kept a cat that became a minor internet celebrity.

But Barlow is better remembered for another cause — letting everyone speak freely and independently online. A Declaration of the Independence of Cyberspace, which he wrote in 1996, is still quoted today.

Governments of the Industrial World, you weary giants of flesh and steel, I come from Cyberspace, the new home of Mind. On behalf of the future, I ask you of the past to leave us alone. You are not welcome among us. You have no sovereignty where we gather.

We have no elected government, nor are we likely to have one, so I address you with no greater authority than that with which liberty itself always speaks. I declare the global social space we are building to be naturally independent of the tyrannies you seek to impose on us. You have no moral right to rule us, nor do you possess any methods of enforcement we have true reason to fear.

The declaration attacked the US government, which had just passed the Telecommunications Reform Act: "This law repudiates your own Constitution and insults the dreams of Jefferson, Washington, Mill, Madison, de Tocqueville, and Brandeis."

Barlow was a committed opponent of copyright, and the declaration may still sound a little extreme today. But it was born in a particular moment: the encryption algorithms that run on every computer and phone now were, back then, controlled by the US government as tightly as weapons.

Beyond the declaration, Barlow's larger influence came through the Electronic Frontier Foundation (EFF), which he helped found.

On the evening of February 7 Beijing time, EFF executive director Cindy Cohn announced on the organization's website that Barlow had died in his sleep, at 70. In her eulogy, Cohn wrote: "It is no exaggeration to say that major parts of the internet we all know and love today exist and thrive because of Barlow's vision and leadership. He always saw the internet as a fundamental place of freedom."

EFF was founded in 1990, when Google, Amazon, Alibaba, Tencent, and Yahoo did not yet exist and the public had no idea what the internet was. Even then, EFF had begun worrying about how government surveillance, encryption technology, leaked personal data, digital copyright, and network firewalls would shape free speech online.

Over the next 28 years, EFF turned up in every kind of digital civil-rights fight. It hired lawyers and won cases that forced the US government to drop export controls on encryption software and to accept that software code is protected speech. EFF has helped jailbreakers fight Apple, and it has gone to court to back Apple against government demands for a "backdoor."

EFF has kept reminding the world what too much concentrated power can do, even in extreme cases that ran against public sentiment. Last year, when internet companies moved to wipe out the far right, EFF and the older civil-rights group ACLU were among the few voices publicly opposing them.

EFF has shaped more than one generation. When Snowden revealed his identity in Hong Kong in 2013, his laptop carried two stickers: one for EFF, the other for Tor, the anonymous communication software EFF funds.

Twenty-eight years on, EFF has won plenty of ground, but the problems its founders worried about at the start have not gone away. In fact, as companies have grown to a scale never seen before and governments have grown more comfortable with technology, many of those problems have gotten worse.

"I discovered that the government had no idea what the internet was" — how the Electronic Frontier Foundation began

The Electronic Frontier Foundation traces its origins to an FBI investigation of Barlow in April 1990.

Barlow was suspected of belonging to the NuPrometheus League, an information-terrorism group that had stolen source code from the memory of Apple Macintosh computers and passed it around.

But Barlow quickly realized that the FBI agent who came to question him had no idea what source code was, and had taken Autodesk — the company behind the architectural design software AutoCAD — for a construction contractor. Before he could clear his own name, he had to turn around and explain to the agent which laws the NuPrometheus League had broken.

From that agent, he saw a much larger problem:

Law enforcement across the United States was confused about the technology, the law, and the very definition of the internet, and in that confusion everyone's online freedom was at risk.

Barlow posted his experience on the computer forum WELL and found that many others had run into the same thing, including Mitch Kapor, founder of Lotus, then the largest independent software company (later bought by IBM). He got in touch with Kapor, and the two began talking about what they could do.

In July 1990, the Electronic Frontier Foundation (EFF) announced its founding. The founders were Barlow, Mitch Kapor, and John Gilmore, an early member of Sun, one of the first commercial computer companies. At the start, Apple co-founder Steve Wozniak and Gilmore both gave EFF substantial funding.

In an interview with The Washington Post, Barlow said EFF's goal was to fight for the rights of ordinary people wherever internet technology and the law collided.

EFF's first case was the Steve Jackson Games affair.

In March 1990, the game company Steve Jackson had founded was suddenly raided by the US Secret Service (USSS), which carried off all of the company's computers and other electronic equipment. The only reason: the Secret Service suspected that Steve Jackson had distributed a government document illegally copied by hackers through the BBS his company ran.

The Secret Service eventually returned all the equipment and decided not to prosecute the company, because it could not find the document. But Steve Jackson discovered that the email stored on his computer had not only been read — large amounts of information had been deleted.

He believed his freedom of speech and his right to privacy had been violated, and he wanted to find an organization that could help him sue the Secret Service for damages. But in an era when the internet was still the preserve of engineers and hackers, no organization understood both the legal weight of these questions and the technology behind them.

EFF then got in touch with Steve Jackson and several other BBS users the Secret Service had investigated, and sued the Secret Service on their behalf. The case took four years, and EFF won: the Secret Service paid Steve Jackson Games $50,000 in damages and $250,000 in legal fees.

More important, it was the first time a US court ruled that email, like a person's phone calls, should be protected from intrusion by privacy law. The case also opened a public argument about how to write laws that protect personal privacy in cyberspace.

EFF won its first case, but that was only the beginning of the conflicts between ordinary internet users and governments and big companies.

EFF once broke the government's grip on encryption, but governments everywhere are back at it

"Encryption may be the only reliable way to protect privacy, but that method is now being restricted by the government. Relying on the government to protect your privacy is like inviting a peeping-tom burglar to install your window blinds."

That comes from an EFF article published in May 1996, about the National Security Agency's export controls on encryption.

At the time, encryption had not yet reached the public or the tech companies, and the US government was already regulating it, barring anyone from disclosing encryption technology to foreigners.

In 1987, the computer of Daniel J. Bernstein, then a student at UC Berkeley, was attacked. Curious about encryption and security, he started working on new encryption methods to keep his machine from being attacked again.

Three years later, Bernstein had written a new encryption and decryption technique based on hash-function transcoding. He then contacted government departments himself to ask whether he could publish it openly. The National Security Agency found him quickly and warned him that if an enemy of the United States in Afghanistan downloaded his paper, Bernstein would likely spend the rest of his life in prison.

Bernstein did not give up. He sued the US government. EFF reached out to him, took the case, and lined up free help from eight lawyers in different places.

Cindy Cohn, the EFF executive director who wrote Barlow's eulogy, was one of those lawyers. At the end of her argument she asked a question:

"Code is a form of speech, and it should be protected as free speech. And is it right for the government to deny a citizen a program that can protect his own privacy?"

A year later, the court ruled that encryption should not be treated only as a state secret — it is also a tool for protecting personal privacy.

That same year, the US government's case against Zimmermann, the computer engineer who had put encryption technology openly online, also ended. After a three-year investigation, he was never convicted. The whole episode came to be known as the Crypto Wars.

After four years and three lawsuits, the US government loosened its restrictions on exporting encryption software. The Ninth Circuit Court of Appeals ruled that software code, as speech, is indeed protected by the First Amendment.

Encryption also began moving into commercial use from that point on. But losing all those cases did not make the government any more open.

In the decades since, friction between governments and tech companies over encrypted data has only grown, and the pressure on the companies has steadily increased.

In 2016, the UK government drafted a national surveillance bill. One clause required telecom service providers, on receiving a lawful interception warrant, to strip the corresponding encryption and hand the data over in a usable form.

According to the Financial Times, British intelligence agencies had already begun notifying American tech companies that they intended to use that clause to demand decrypted data.

In the United States, twenty years after the Crypto Wars, the standoff between encryption and government flared up again in 2013 over Snowden.

In 2004, Apple, which kept upgrading the encryption on its smartphones, hit a deadlock with the US Department of Justice: in an investigation involving guns and drugs, Apple was asked to hand over text messages from a suspect's iPhone, and Apple said iMessage was encrypted and it could not help.

In 2016, Apple published an open letter refusing a US government demand. The government wanted Apple to unlock an iPhone used by a criminal suspect.

What the FBI asked for was a special operating system with two features:

• after 10 wrong passcode attempts, the system would not erase the data;
• the passcode could be entered over a cable rather than by finger on the screen, so thousands of guesses could be tried quickly.

Put plainly, with a system like that, the FBI would not need to break the encrypted data at all. It could break the iPhone's lock screen and take everything.

Tim Cook refused, and nearly every tech company in Silicon Valley backed his decision to one degree or another.

The reasoning was simple: if every tech company were forced to install a government-specified backdoor, how could any of them offer users a service worth trusting?

In Cook's words: "In the physical world, it would be the equivalent of a master key, capable of opening hundreds of millions of locks — from restaurants and banks to stores and homes. No reasonable person would find that acceptable."

But the company that publicly refused the US government was, after all, one of the largest tech companies in the world, with many other tech companies speaking up behind it.

What about the smaller ones? Could they afford a long and hugely expensive lawsuit against the US government?

Lavabit, the encrypted email service that hosted Snowden's account, refused to hand the government its decryption keys. The small company ended its standoff by shutting down.

As the internet has spread, phones, computers, and all kinds of electronics have taken over our lives. More and more of our information and our private lives have turned into data made of 0s and 1s.

Without encryption, everywhere you go, everything you say, everything you do can be watched. Compared with 20 years ago, the government can learn far more about you than it used to.

No one had to invite him. The man with the window blinds is already outside.

The internet is the freest tool humans have ever invented, and also the best surveillance tool

In June 2013, Snowden sent several thousand classified documents to three journalists. They showed that the National Security Agency (NSA) had been tapping phone calls and monitoring the email and web activity of millions of Americans and of citizens of other countries, and they set off a global argument.

Over the following year, more and more evidence came out, and photographs of Snowden began appearing in the press — along with the conspicuous red EFF sticker on his laptop. While Snowden was on the run in Hong Kong, EFF kept calling on the United States to protect him.

A reporter asked John Perry Barlow at the time what he made of the PRISM program. His answer:

"The first words that come to mind are: I told you so! For years EFF has been warning about the NSA's surveillance operations."

As far back as 2000, EFF published a document the NSA had submitted. In it the agency said that the huge volume of data the internet produced was making it extremely hard to find intelligence and protect national security, and that the NSA had to "live on the network" at all times.

A year later came 9/11. The United States stepped up its fight against terrorism. In October 2001, President George W. Bush signed the USA PATRIOT Act, whose Section 215 said that, to counter international terrorism and espionage, the government could inspect any tangible records.

The NSA and other agencies read that clause to mean that monitoring every phone call and email record inside the United States was "entirely legitimate."

By 2004, EFF had discovered that NSA staff had been working inside the American telephone carrier AT&T for a long time. In January 2006, EFF filed a class-action suit against AT&T, saying the company had allowed and assisted the NSA in illegally monitoring users' communications inside the United States.

The case was dismissed and appealed several times. In 2012, the US Supreme Court declined to hear the appeal, which meant EFF had lost. But the EFF lawyers on the case said they would keep exposing more about state surveillance.

After PRISM broke, EFF launched more campaigns against mass state surveillance.

In 2015, EFF published a plan to end mass surveillance: it urged tech companies to strengthen encryption in their systems to keep the NSA out, encouraged users to protect their own information with encryption software, called for more secure communication apps and tools, and pushed for legal principles on oversight and privacy.

EFF also produced a guide to help internet users protect themselves from surveillance, covering how to use encryption software, how to store private material properly, and how to assess your own risk of being watched.

Mass state surveillance is not only an American matter. In 2007, the group Privacy International studied 48 countries. Eight of them — including China, Malaysia, Russia, the United States, and the United Kingdom — were rated "endemic surveillance societies."

In September 1998, construction began on the Golden Shield project, which called for a central hub to monitor and manage China's internet. The system later grew to include network monitoring, video surveillance, and facial recognition.

In November 2000, at the International Exhibition on Public Security Products and Police Equipment 2000 in Beijing, the scope of the Golden Shield project was made public: monitoring entry and exit points in public places, decrypting and encrypting information, e-commerce security, communications security. No one explained where the data the project collected ended up.

In 2006, the first phase of Golden Shield passed state acceptance; the second phase was finished in 2008.

If the arrival of the internet changed the relationship between people, it was bound to change the relationship between governments and ordinary citizens too.

From knowing nothing about the internet to building an enormous automated surveillance network, mass state surveillance has climbed along with the internet. The protection of personal privacy has not kept pace.

"The internet is the freest tool humans have ever invented, and also the best surveillance tool. It isn't a choice between freedom and surveillance — it's both." Barlow scribbled that on a page of courtroom transcript in 1998, and it was later quoted in an article on EFF's website.

Digital privacy keeps widening, and people seem to have less and less control over it

In 1990, after EFF helped Steve Jackson win his case, email was brought under the protection of privacy law.

The United States had already passed the Electronic Communications Privacy Act (ECPA) in 1986. But at the time that only extended the existing rules on telephone wiretapping into the world of computers. How far and how strictly email, online forum data, and companies' use of users' private data were governed remained unclear.

In 1991, EFF began working with Computer Professionals for Social Responsibility (CPSR), and over two years drew up a set of policy proposals for protecting computer and citizen data privacy. The documents aimed to keep watch on how far governments and companies could intrude on private life; an ISP intercepting information, for instance, would not be allowed. EFF also planned to provide information to organizations and individuals interested in citizens' privacy.

In 1999, EFF ran a "12 ways to protect your privacy" campaign, using a guide to show people the dangers of leaked personal data — including checking big companies' privacy terms and making sure the contents of an email give nothing private away.

EFF later wrote a report called On Locational Privacy, and How to Avoid Losing Yours. Over the next decade, EFF said, digital systems woven through every kind of public space would record and store where people went; such systems were already appearing, and more were coming.

EFF put particular weight on intrusions into personal location data. Location information, it said, was being quietly collected by devices everywhere, and it could be analyzed, searched, and even sold.

Technology's intrusion into privacy is no longer a new subject. It is just that the spread of the mobile internet has changed what privacy, acquaintance, stranger, and public space even mean.

Twenty-some years on, digital privacy has long since expanded from email to every trace you leave in the digital world.

Smartphones have worked their way into people's lives more thoroughly than any device before them; your phone understands your life far better than other people do. It can give your private information away in ways you would never expect.

Websites that carry advertising often run more than half a dozen tracking scripts. A smartphone camera usually attaches location data automatically. Map apps know where your home and your office are, and when you leave work. Search engines know things about you that you may not know yourself. Together they build a complete portrait of the user.

Meanwhile governments and intelligence agencies are both working with tech companies and working against them. The Snowden documents, which showed the NSA colluding secretly with some tech companies while hacking others to steal user data, were only the tip of the iceberg. Government agencies have never stopped pressing Apple and other companies to build "backdoor" technology that gets around strong security.

EFF has given money to the dark-web browser Tor. Tor began as a project at the US Naval Research Laboratory, meant to protect American online intelligence communications. The name stands for "The Onion Router." In 2004 the laboratory cut the project and released the code. EFF funded it; with Tor, internet users can post completely anonymously and hide their tracks.

In 2014, EFF released Privacy Badger, a privacy extension for Chrome and Firefox that blocks tracking by advertisers who ignore user privacy agreements. By April 2017 it had 1 million users.

To sell ads better, the big tech companies keep pushing past the line on user privacy, and the arguments over how free internet services should be allowed to use personal data keep multiplying. But most people simply hand their digital privacy over without a thought.

Digital copyright has found its footing, but the subscription model raises a new question of ownership

Digital copyright infringement is an old subject. In the 20-some years since the Digital Millennium Copyright Act of 1996, companies in music, publishing, and games have gradually built copyright walls with technical means, and being "legitimate" has come to look more like a weapon in the fierce competition between companies.

From its founding, EFF paid attention to copyright in electronic content. In 1997, EFF published an article on its site called "How to Think About Copyright on the Internet," in which co-founder Mitch Kapor discussed a contradiction: the internet spreading traditional content like printed books and poetry collections for free and instantly, set against copyright.

"For the copyright of digital content, the new internet era needs a different kind of solution," he wrote at the end.

But EFF was already thinking further ahead, because copyright protection had created new problems of its own.

In 2009, because of a change in copyright law, Amazon's Kindle store pulled George Orwell's novels 1984 and Animal Farm. It was a thorough removal: through remote syncing, Amazon also wiped the copies customers had already bought and downloaded to their readers.

The irony is hard to miss — 1984 itself describes something very close, with the Ministry of Truth recalling published newspapers to destroy what Big Brother had said earlier.

Bruce Schneier, then chief security technology officer at British Telecom, said: "When you buy a book from Amazon, you get shockingly few rights… I can't lend the book to someone else, I can't resell it after I've read it, and now it turns out I don't even know whether my book will still be there tomorrow."

EFF lawyer Fred Von Lohmann said the episode made people see for the first time the gulf between assets in the virtual world and the real one: "There's a world of difference between buying a physical book and buying an e-reader, but consumers hold the same expectations of both."

Under Amazon's user agreement at the time, an e-book downloaded to a reader was not yours. The remote deletion was rather like buying a book from a bookstore and having the store send someone a few days later to pick your lock and take it back.

Amazon CEO Jeff Bezos apologized publicly afterward, calling the company's solution "stupid, thoughtless, and painfully out of line with our principles." Nothing like it has happened since — books still get pulled, especially from the Chinese store — but Amazon no longer deletes them remotely.

Not every company works that way, though. Elsewhere, the right to use a digital asset has been stripped away from owning it even more completely.

In 2008, Spotify launched with licensed music streaming and opened up a new subscription business: users could pay $9.99 a month for unlimited access to and downloads from the whole catalogue. You no longer had to download music to your computer or phone; it all sat in the cloud, there whenever you went online.

Streaming was cheaper and more convenient, and the subscription model rescued a record industry sliding downhill. Then came Netflix, selling video on the same monthly plan. The model is familiar by now: NetEase, Tencent, Xiami, iQiyi — every mainstream video and music service is pushing its own paid membership.

And a new problem arrives. Digital content has now come loose from anything physical at all, even the file you used to download onto a device. The e-book is not on a shelf, the CD is not in a record store; if you insist on finding the physical thing, it is on a server in a city you have probably never visited, and the file downloaded to your phone cannot be opened any other way.

This kind of consumption looks much more like renting than buying. Every month you pay an internet company a fee for the right to use a virtual good.

Remote deletion has become easier, and harder to notice.

Starting in January last year, NetEase Cloud Music, QQ Music, and Xiami Music removed, one after another, the entire catalogues of singers including Chen Sheng, Vivian Hsu, and Anthony Wong. On October 19 last year, Apple Music's mainland China catalogue deleted everything by the Taiwanese singer Chen Sheng, including music users had already paid to download. And under the terms of service, the platform bears no responsibility at all.

More countries are starting to fence the flow of information along national borders

"Cyberspace does not lie within your borders. Do not think that you can build it, as though it were a public construction project," Barlow wrote in the 1996 Declaration of the Independence of Cyberspace, describing the limits being placed on the online world.

EFF pointed out at the time that in China, Germany, France, Russia, Singapore, Italy, and the United States there were signs that governments were trying to block the spread of information online by building network firewalls.

Back then Chinese internet users could still reach overseas websites normally. What put EFF on guard was that in February 1996 China issued its first regulation on the internet, the Interim Provisions on the Administration of International Networking of Computer Information Networks of the People's Republic of China. It required that network access run through the international gateway channels provided by the state public telecom network, and it called for tighter management of information online.

That was also when the national firewall began to be designed. Fang Binxing, the former president of Beijing University of Posts and Telecommunications, said in a 2011 English-language article in the Global Times that by 1998 the national firewall was nearly built.

In 1995, Russia passed the Federal Law on Information, Informatization and Protection of Information, which drew a clear line around what information was open and what was secret. The Russian Federal Security Council then set up an information security division to coordinate and lead the country's information security work.

Four years later Russia widened its oversight of the internet. The Federal Security Service, the Ministry of Internal Affairs, and the federal media and culture agency all set up network departments to monitor state information, news media, and hacker forums. The Interior Ministry's network monitoring center was devoted to watching social media such as Twitter and Facebook.

As the internet spread after 2000, more and more signs showed that EFF's old worry about borders online was not unfounded.

Today China, Russia, Brazil, and other countries all have explicit rules that citizens' data must not leave the national network.

Russia and Brazil both introduced such rules in 2016, strictly requiring data to be stored at home. Leonid Levin, deputy chairman of the Russian State Duma's information policy committee, said at the time: "A large amount of Russians' personal data online is stored on foreign servers, especially American ones. There's no way to check whether it contains information that threatens national security."

But whatever the stated purpose of the checking, the real aim is to keep personal data somewhere controllable.

For big tech companies serving customers worldwide, this is not easy to accept.

At first, multinationals including Apple and Microsoft tried to resist the drawing of borders across the internet and the censorship of information. In their biggest markets, they did not hold out long.

In November 2016, the National People's Congress passed the Cybersecurity Law, which took effect last June. Article 37 requires: "Personal information and important data gathered and produced by critical information infrastructure operators during operations within the territory of the People's Republic of China shall be stored within that territory."

The rule means every internet service operating in China must be stored physically inside China, and must go through the public network to connect abroad.

Apple ended up moving iCloud into China. Last July it announced a $1 billion investment in a new data center in Gui'an, Guizhou, to run iCloud services there.

With iCloud in China, Apple's latest iCloud terms of service say: "You understand and agree that Apple and GCBD [Guizhou-Cloud Big Data] have access to all data that you store on this service."

Faced with Russia's 2016 data security law, LinkedIn — unwilling to hand over its Russian users' personal data — sued Russia's national information security agency instead.

In November 2016 a Russian court rejected the appeal, insisting that unless LinkedIn changed its mind and stored the data on servers inside Russia, it would leave the country for good.

By the end of that November, internet providers inside Russia had begun blocking LinkedIn's IP addresses.

John Perry Barlow once said in an interview: "Governments will always refuse technological progress they can't control. With something like the internet, they fear the possibilities that open up when the barriers to data flow come down."

Now governments in country after country are building more barriers to slow that flow.

A few companies hold the technology and the traffic — which means they can control speech too

On the internet, everyone has freedom of speech. Barlow's 1996 Declaration of the Independence of Cyberspace kept coming back to that point.

Protecting free speech online was part of EFF from the start. In 2001, a federal court in Washington ordered the internet service provider TheMart to disclose the identities of users who had posted anonymously on the company's BBS, because the US government wanted to stop what it saw as their inflammatory discussion of certain policies.

EFF represented TheMart in challenging the order. Seven months later a Seattle superior court declared the subpoena for user information invalid, and held that users' freedom of speech should be protected.

"This is an important ruling for free speech on the internet. The court recognized that you should be able to express your views online without fear that your speech will be controlled by an unrelated lawsuit," said Aaron Caplan, the EFF lawyer on the case.

Governments can regulate speech, but the companies that supply the internet's "infrastructure" have a more direct kind of power.

That argument reached its peak last year over a website called Daily Stormer.

In August 2017 a demonstration broke out in Virginia, set off by the town of Charlottesville's decision to take down a statue of General Lee. The famous Civil War general stood for the defense of slavery, and local white supremacist groups held a rally to protest the removal.

Opponents gathered too, and on August 12 the two sides clashed. A dark gray Dodge sedan drove into the crowd demonstrating against the far right, killing Heather Heyer, a 32-year-old white woman, on the spot.

The next day the far-right site Daily Stormer published an article calling the dead woman a "childless slut." That is classic Nazi rhetoric; the site takes its name from Der Stürmer, a Nazi publication from the Second World War.

Daily Stormer became a target at once, and the internet companies began cutting it off.

At noon the day after the article appeared, GoDaddy, the world's largest domain registrar, said it would no longer serve Daily Stormer and gave it 24 hours to find another host. The reason GoDaddy gave was that the site had "violated the terms of service."

Without a domain, a website does not exist for the vast majority of internet users.

Hours before GoDaddy's statement, Daily Stormer had already applied to move to Google's domain registration service. In less than three hours, Google refused for the same reason. Google also shut down Daily Stormer's YouTube account while it was at it.

After some scrambling between domain hosts in France and Russia, Daily Stormer ended up with no domain of its own, no accounts on any major social network, and no presence in Google's search results.

No government order and no coalition demand made the internet companies do this. Daily Stormer simply disappeared.

The speed and force of it worried EFF. EFF is not an organization that supports Nazis, but it published an article condemning Silicon Valley's collective silencing of right-wing speech.

"We protect free speech not because we support one kind of speech or think that kind of speech deserves support, but because we believe that neither governments nor commercial organizations have the authority to interfere with free speech."

EFF argued that the tools used to block hate speech will eventually turn back on the civil-rights organizations fighting for freedom, and that the big tech companies will one day use them to silence those groups.

What EFF feared is getting worse. Silicon Valley, which once treated liberalism as gospel, used to tolerate extreme speech fairly broadly. The mood has changed in the past few years.

Beyond domain names, an internet company with a large user base can wipe out an entire strain of speech simply by changing what it shows people.

After the killing, Facebook deleted links to Daily Stormer's articles and its event pages; Twitter cleared out the organizers' accounts; Reddit closed a forum full of related violent rhetoric. Between them, those three platforms cover nearly every channel for speech online in the United States.

Even the payment companies PayPal and Apple Pay stopped providing payment and transfer services to extremist groups.

Daily Stormer is an extreme case. More often, bias forms out of sight, inside the daily algorithmic recommendations of the big platforms.

Facebook's Mark Zuckerberg said his goal for 2018 was to adjust the algorithm to show users less content from public sources such as news media, and to give priority to posts and comments from family and friends. Over the past year Facebook has taken heavy criticism over fake news.

Facebook is not the only one deciding who gets to see what, and certainly not only in the United States.

Cover image from Variety, asgw on Visualhunt / CC BY