The Worldwide Web Of Walls

Al Gore didn’t invent the Internet. He did, however, articulate one of the strongest visions of its promise.

In March of 1994, the vice president flew from Washington to La Paz for seven hours, where he signed accords with Bolivian leaders to strengthen the two countries’ cooperation in fighting the narcotics trade. From there he traveled to Buenos Aires to address a conference of global telecom experts with a homily about a world in which people could communicate on equal terms no matter where they lived.

“We now have at hand the technological breakthroughs and economic means to bring all the communities of the world together,” Gore told the assembled audience of International Telecommunications Union members. “It will be a means by which families and friends will transcend the barriers of time and distance.”

A generation later, the global community Gore imagined is divided by its own internal barriers. If Gore followed the same itinerary today, he might find himself flying across several distinct Internets, all vastly different from the one he and his family would access at home in the United States — with abrupt changes at national borders. Meta’s recent commitment, part of a legal settlement with state attorneys general, to impose time limits and “nighttime blocks” on Facebook and Instagram scrolling time, applies only to U.S. users between 13 and 17. At the same time, lawmakers in some Mexican states are considering their own social-media age restrictions for users under 16. In Brazil, a federal electoral court limits AI use by political campaigns; in neighboring Argentina, where Gore gave his speech, a strong national data privacy law protects consumers but also makes it more complicated for certain online industries (like crypto) and social media companies (including Meta) to operate.

While human rights monitors have trained their attention on how authoritarian governments have constrained Internet access for their citizens, the idea of a “global Internet” has failed for reasons much broader than firewalls in China or Cuba. Instead the fragmentation now comes largely from democratic societies enacting their own rules to prioritize privacy and child safety, defend public order and police disinformation. The growing distrust of tech companies in democratic countries has led to an age in which their governments are the ones that are the most active in pushing new legislation that constrains their citizens’ ability to access parts of the Internet.

The result is a patchwork of online regulations that are quickly becoming more complex, more strict and harder to reconcile from one jurisdiction to the next. Without a shared technical framework and the persistence of local differences, the contrast between what happens online across national borders will only get more dramatic.

“There’s a bit of a myth that the internet has always been equal for all of us,” said Jillian York, an American-born free speech activist and the current Director for International Freedom of Expression at the London-based Electronic Frontier Foundation. “Those of us from more democratic societies are waking up to a reality that the rest of the world has already experienced.”

Welcome to the age of the isolationist Internet.

The modern Internet was one year old when Gore landed in Buenos Aires. In 1993, Swiss-based intergovernmental research institute CERN put the Internet into the global domain — meaning the core technology would be free and open for all to use. It augured the end for “walled gardens” like those maintained by dial-up platforms like America Online or dedicated hardware like France’s Minitel. But users wanted an open Internet, and soon people around the world were using novel web-browing tools to navigate the quickly expanding network.

No single company or body could control who built or accessed a website, which fed a utopian vision: all of a sudden, everyone around the world was using the same tools to access a shared pool of information. In a world where everything from electrical outlets to health-care systems were different, the Internet and its associated instruments were remarkably alike — the globe shared browsers, social networks, apps, URL architecture and even coding languages. Gore’s speech captured the optimism of an era in which western political leaders hoped to see goods and services, but also ideas, moving freely around the globe with the Internet as their conduit.

The greatest impediment to Gore’s idealism appeared to be autocratic governments. Many constructed complex digital gatekeepers to control incoming and ongoing web traffic and throttle their citizens’ access to the information, deploying new technology as part of broader mechanisms of political control. As early as 1997, digital observers were referring to the “Great Firewall of China,” but America’s leaders believed its crumble was inevitable.

“China has been trying to crack down on the Internet,” Bill Clinton said in a 2000 speech. “Good luck! That’s sort of like trying to nail Jell-O to the wall.”

It was not just totalitarians unwilling to let their citizens roam freely online. Other governments began blocking websites, throttling traffic and punishing those who used the Internet to access or share content already taboo or banned in other media. In 1996, Singapore extended existing broadcast restrictions to cover online content “against public interest or order, national harmony or which offends against good taste or decency.” That same year Germany shut down access to foreign neo-Nazi sites that prosecutors said were “helping to incite racial hatred.”

Gore imagined that these sort of site blockages and national differences could slowly melt away, and Clinton’s choice of an American corporate trademark for his simile was no accident. American leaders believed a globalized world would lead to a prosperous and democratic one in which the digital gatekeepers could not survive. Authoritarians would have little ability to regulate their citizens in an Internet-connected world, which would then make their entire system of governance obsolete.

They failed to account for the way in which free access to the Internet would make many people turn against it — specific tools and platforms, the companies behind them, even the idea of technologically-centric human progress. Public opinion has shifted against big tech, driven specifically by a decline in trust over the past decade in Western markets. There is a similar decline in trust of AI systems that follows that pattern: Consumers in the developed world have less trust of AI than those in the developing world. Citizens fear disinformation, addiction, energy use, hate speech and online platforms’ radicalizing power — and governments have responded.

In fact, it is now democracies rather than authoritarian régimes that are imposing the most creative new controls on Internet traffic. Legislatures have drafted new laws, executives have interpreted their existing authority to tighten access, and lawsuits against social-media companies have garnered significant settlements because companies fear juries will rule against them.

Each country is going alone, with their own rules and on their own timelines. A law in Turkey that took effect in 2020 aims to force social media companies to appoint local representatives responsible for content, store data within the country and respond to content removal complaints within 48 hours. Earlier this year, India temporarily restricted the messaging app Telegram in an effort to cut down on fraud related to medical school entrance exams. Tanzania, Uganda and Kenya have all recently imposed various content licensing requirements that differ slightly from one another and from other countries’ licensing regimes.

Governments imposing many of these restrictions are doing so not over the complaints of their citizens but at their behest. As anger at tech companies grows and takes distinct forms around the world, national governments begin to tailor their responses.

“The main story is that we have a growing global understanding, an emerging consensus-like approach that these regulations are necessary,” says Anu Bradford, the Director of the European Legal Studies Center at Columbia University, speaking mostly about age-limits for children but also some other forms of regulation. But, she adds, “there are more players, more regulators and less trust between players.”

Sometimes subnational units (like California, where Gov. Gavin Newsom signed a bill banning teens from accessing allegedly addictive features of social media apps on September 10) and supranational bureaucracies (like the European Union, whose General Data Protection Regulation imposes harsh standards on how companies handle users’ information) are getting ahead of nation-states. The settlement of the Meta lawsuit, brought by 29 U.S. states, has left the social-media behemoth cajoling its competitors to accept many of the same regulations it was forced to adopt, in an effort to save face and keep pace.

Even when governments have not imposed clear restrictions, they are shaping distinctly different online experiences. In the name of “digital sovereignty” the French government is imposing its own Zoom alternative, while other European governments are trying to push their populations — often starting with government officials — off other U.S. software.

But even as debates flare up in national capitals about where to draw the line between a child and an adult — debates informed by local cultural expectations and other rules around voting age or drinking age — any effort to actually enforce online age limits will immediately impact everyone in the country who goes online. Attempts to confirm users’ ages have taken different forms in different governments, but the most strict regimes have installed a whole new privacy apparatus, forcing everyone regardless of their age to verify their identity before using certain platforms.

“I don’t think it’s yet particularly well understood what the full implications could be in the long run,” says Vicki Nash of the Oxford Internet Institute in the United Kingdom, referring to a 2023 law that required broad age verification, content removal and safety risk assessments of content that children were seeing online. “When the legislation around child safety was introduced, there was an understanding that companies would need to know who is a child. There hasn’t been much debate about how you’re identifying who is a child.”

The patchwork of regulation, different and ever-changing from country to country, also adds significant regulatory burdens to private companies. Startups, even when fairly well-funded, are often unable to afford the lawyers in every country necessary to untangle each specific set of laws — and build different iterations of their products that comply with each. This means some of these younger companies not begun or funded by existing tech giants will often choose to not operate there at al — ironically sometimes adding to the power of big tech that some of these regulations are meant to curb.

For users, that means less choice — the tech giants pay up to operate around regulations, while newer companies are forced to think smaller in their ambition. It results in the web stopping at the border: Each country subject to an odd pastiche of national regulations of online behavior that are rapidly being deployed.

Imagine you are a Portuguese 14-year-old, heading from Lisbon with your parents on a summer train trip that will end in Copenhagen, where some old relatives are waiting to greet you. Eurail, for all of its benefits, doesn’t connect well to Madrid, so you’ll take a puddle-jumper flight and begin your train journey in Madrid’s grand Atocha railway station, home to quite a lot of flora and fauna, not to mention old Ben Lerner novels.

Your parents, unfortunately, are strict. In your home country, you need parental consent to access social media between the ages of 13 and 15; they don’t quite trust you yet. But when you land in Spain, you’re allowed to open a social media account starting at 14 (though a proposed law would take that age up to 16). Unbeknownst to your parents, who for some reason are working themselves into a tizzy about where your train boards from even though it does not leave for two hours, you download Instagram and make an account. You start scrolling and pretty quickly are enjoying your time on the app.

But now, your train departs for Marseille. And as you cross into the South of France, you were prepared to log off — one of the French kids you know from streaming video games on Twitch had warned you that the parliament had passed a social-media ban for children under the age of 15. Hold on, though, because the French Constitutional Council just struck down the law, arguing it disproportionately infringed on minors’ right to freedom of expression and communication. You’re safe to scroll, for now. Still, France’s digital regulator Arcom is very powerful and effective at stopping piracy. The sorts of content that you might encounter on social media (regardless of age) are therefore different — there are probably fewer unlicensed movie clips that appear on your algorithmic feed, for example.

After a night in a shabby Marseille hotel, you head to Zurich via Italy. During your layover in Milan, all you really have to worry about is pornography-related age verification. As a horny but also rule-following teenager, you encounter this system and quickly decide not to push your luck trying to get around it.

Once you pass into Switzerland, you are perceptive enough to pick up one small difference in the browsing experience. The familiar “Accept all” and “Reject all” choice that pops up in a banner before a website places new cookies in your browser, a function of European privacy requirements, has been replaced by Switzerland’s own digital privacy network which generally includes an “opt out” button for cookies; the country sits beyond the EU’s digital regulatory system.

As you cross from Switzerland into Germany towards Frankfurt, you might not notice any obvious differences to your algorithm. But they are there. Germany has extremely stringent regulations around hate speech online. “Incitement to hatred,” which includes actions like denying the Holocaust online, is punishable by up to five years in prison. These regulations both filter out this sort of content and serve as a deterrent against this sort of posting in the first place. The internet that you encounter here feels different — there’s less inflammatory content that is meant to draw anger and engagement — though the toxic sludge still seeps through.

Finally, you have reached Copenhagen, and the cousins you see once a year. You’re less invested in your phone because you actually want to hang out with them, good luck for you because Denmark is in the midst of approving a law to restrict social media use for children under 15 without parental consent, though it’s not being enforced quite yet. But when you try to use your uncle’s old desktop computer to upload a week’s worth of digital photos to your family’s shared OneDrive cloud storage account (due to their weird vendetta against paying for iCloud storage) you are unable to log on — as a Danish civil servant the government has been pushing him to abandon Microsoft products and use open-source LibreOffice instead.

You passed through seven countries on your trip. As an EU citizen, your passport wasn’t stamped once. You used the same money nearly everywhere, your phone plan’s minutes-and-data allowance translated effortlessly across borders and the food you ate was made under common standards and shared the same labeling style. But as soon as you went online, the grand experiment in cross-border interconnectedness ended.

The dream of the 1990s, as articulated by Gore, was the manifestation of an even older idea — that of the “global village.” Coined by Canadian media theorist Marshall McLuhan in 1962, the concept evolved to be that of the world becoming increasingly interconnected due to the advent of media technologies.

McLuhan is often referred to as the “man who predicted the Internet.” But his idea of the “electronic age” includes a move from the culture of individuality towards a collective identity brought on by digital technology. The internet did not flatten national interests. Instead, what Bradford calls the “geopoliticization of regulation” has inspired a new set of national powers to manage a technology that inherently helps people connect across borders.

A contemporary of McLuhan’s, political scientist Benedict Anderson, was not a media scholar but did theorize a concept about national identities — namely, that they had been constructed in large part due to an increase in mass literacy. Anderson called them “imagined communities.” These “imagined” tribal identities that Anderson discusses have only been reinforced by the Internet age. Even as everyone has more information than ever about the rest of the world, they still gravitate towards and hold tight to national character.

These norms are both built and reinforced by the regulations and policy choices that shape distinct, national Internets. In this way, the isolationist Internet is recursive. It is shaped by regulations that are themselves reflections of culture and identity — distinct values about privacy, free expression or what it means to be an adult — and then it helps to maintain and strengthen those national differences. The regulatory environment builds on itself.

And as local politics become more tribal, more dependent on questions of nationalism, the Internet both reflects this phenomenon and deepens it. The world, as it turns out, never wanted to see its villages globalized.