On a cold night in February of 1996, John Perry Barlow found himself at a party in Davos. It was the closing event of the World Economic Forum, and the ballroom was filled with besuited masters of the universe and students from the University of Geneva. He danced with them, a little inebriated. But a thought nagged at him.
Earlier that day, in Washington, D.C., President Clinton had signed a bill that would for the first time bring the Internet under a degree of government control. The Communications Decency Act (C.D.A.), part of the Telecommunications Act of 1996, included a provision that would criminalize “obscene” or “indecent” content on the Internet. In Congress, the Nebraska senator James Exon, who had co-sponsored the C.D.A., issued a dire warning: “Barbarian pornographers are at the gate, and they are using the Internet to gain access to the youth of America.” As evidence, he circulated a blue binder filled with pornographic material collected online, including an image of a man having sex with a German shepherd.
Barlow, a former cattle rancher from Wyoming, a sometime lyricist for the Grateful Dead, and a libertarian activist on the Internet, was convinced that the fledgling network should remain free of government interference. Incensed by what he would call a “stunningly dumb bit of legislation,” he set up a makeshift office adjacent to the party and, shuttling back and forth between his computer and the ballroom, banged out an eight-hundred-and-fifty-word manifesto. Barlow’s “A Declaration of the Independence of Cyberspace” would soon—in a term that gained currency only later—go viral. It is now recognized as a seminal document in the history of the Internet: a preamble to a constitution that the network would never formally have.
“Governments of the Industrial World, you weary giants of flesh and steel, I come from Cyberspace, the new home of Mind,” the manifesto began. “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.”
Like so many constitutional provisions these days, Barlow’s “Declaration” has recently come under considerable strain. Critics denounce it as an exemplar of techno-utopianism, enabling the uncontrolled, mob-fuelled Internet we have today. The years have not proved kind to Barlow’s vision of “a civilization of the Mind,” more “humane and fair.” Amid the scandals concerning privacy, misinformation, polarization, threats to teen-age mental health, and even complicity in genocide, the radiant future that Barlow foresaw has given way to what the activist and writer Cory Doctorow calls the “enshittification” of the Internet.
In fact, for all Barlow’s outrage, governments remained mostly hands-off during the Internet’s early history. Clinton may have signed the C.D.A., but his real attitude was summed up by his statement that regulating the Internet was like “trying to nail Jello to the wall.” Large parts of the C.D.A. were later invalidated by the Supreme Court on First Amendment grounds, and buried within the act itself was a clause that has over the years come to emblematize the long leash granted to the Internet: Section 230 of the act protects online platforms from liability for content created by their users.
During the past decade or so, however, governments around the world have grown impatient with the notion of Internet autarky. A trickle of halfhearted interventions has built into what the legal scholar Anu Bradford calls a “cascade of regulation.” In “Digital Empires” (Oxford), her comprehensive and insightful book on global Internet policy, she describes a series of skirmishes—between regulators and companies, and among regulators themselves—whose outcomes will “shape the future ethos of the digital society and define the soul of the digital economy.”
Other recent books echo this sense of the network as being at a critical juncture. Tom Wheeler, a former chairman of the F.C.C., argues in “Techlash: Who Makes the Rules in the Digital Gilded Age?” (Brookings) that we are at “a legacy moment for this generation to determine whether, and how, it will assert the public interest in the new digital environment.” In “The Internet Con” (Verso), Doctorow makes a passionate case for “relief from manipulation, high-handed moderation, surveillance, price-gouging, disgusting or misleading algorithmic suggestions”; he argues that it is time to “dismantle Big Tech’s control over our digital lives and devolve control to the people.” In “Read Write Own” (Random House), Chris Dixon, a venture capitalist, says that a network dominated by a handful of private interests “is neither the internet I want to see nor the world I wish to live in.” He writes, “Think about how much of your life you live online, how much of your identity resides there. . . . Whom do you want in control of that world?”
Questions of control have always hovered over the Internet. Its decentralized architecture has long been key to its identity, wielded as a form of originalist rhetoric against any suggestion of external intervention. The roots of this architecture are, in fact, somewhat murky—attributed, variously, to an effort at sharing computing resources more efficiently, a nineteen-sixties confluence of technocracy and hippie anarchism, and the search for a network design that could withstand nuclear attack (a claim disputed by some Internet veterans). In the 1999 memoir “Weaving the Web,” Tim Berners-Lee, often called the father of the World Wide Web, likened the network’s principles to those upheld by his Unitarian Universalist church—individualism, peer-to-peer relationships, “philosophies that allow decentralized systems.”
In truth, the notion of a fully decentralized network has always been something of a myth. The Internet Corporation for Assigned Names and Numbers (ICANN), which has been described as the “secret government of the Internet,” has long managed a directory—the Domain Name System, or D.N.S.—that the Internet needs in order to function. (For Berners-Lee, the D.N.S. was a “centralized Achilles’ heel” that could bring the network down.) Until 2016, ICANN was under the authority of the U.S. Department of Commerce. In a 2006 book titled “Who Controls the Internet?,” the law professors Jack Goldsmith and Tim Wu described “the death of the dream of self-governing cyber-communities,” and argued that governments had an array of means at their disposal with which to enforce their laws in cyberspace, even if imperfectly.
In retrospect, the real problem with the cyberspace-sovereignty argument was simply that it was blinkered. Early Internet activists like Barlow were so focussed on the risks of government intervention that they failed to anticipate the threats posed by private-sector control. This was perhaps unsurprising. Barlow was writing amid the end-of-history glow produced by the collapse of Communism, his techno-utopianism a variation of the era’s market utopianism. The mood has shifted considerably since then. Today’s digital activists came of age in the shadow of 2008; they tend to call for government intervention, in order to rescue the Internet from what the former Greek finance minister Yanis Varoufakis calls “technofeudalism,” in a book by that title.
The dramatic rise of generative artificial intelligence has only accelerated calls for government intervention—and, significantly, these calls are often coming from within the industry. Sam Altman, the recently reinstated head of OpenAI, went before Congress last spring and essentially demanded regulation; Elon Musk has called for a federal department of A.I. In “The Coming Wave” (Crown), Mustafa Suleyman, a co-founder of DeepMind and of Inflection, two leading A.I. companies, argues that government intervention is necessary to protect us from the technology’s enormous risks. (“At some stage, in some form, something, somewhere, will fail,” he writes, in what’s generally a judicious account. “And this won’t be a Bhopal or even a Chernobyl; it will unfold on a worldwide scale.”)
Activists have every reason to hope that A.I. anxieties will bolster their efforts at Internet governance. Yet they’re so attuned to the difficulties of the present that their remedies may do little to nurture a broader set of values—freedom, solidarity, equitable access to resources—that the Internet once promised to advance. The perils of the libertarian approach are now clear; we may soon be learning the costs of reflexive statism. More than a thousand A.I. policy initiatives across sixty-nine countries have lately been documented. In the U.S., some thirty states are debating (or have already enacted) digital-privacy bills, adding to federal oversight by agencies such as the F.T.C. and the S.E.C.
“Look, Dave, I can see you’re really upset about this,” Hal, the digital brain in Stanley Kubrick’s “2001: A Space Odyssey,” tells his human minder, in a tone that calls to mind the bland neutrality of today’s chatbots. “I honestly think you ought to sit down calmly, take a stress pill, and think things over.” In the movie, Dave is right to fear the worst. Amid the rush to regulate, though, Hal’s advice might be worth taking.
In “American Capitalism” (1952), the first volume in a trilogy on economics, John Kenneth Galbraith outlined his notion of “countervailing power.” He was living in a time—much like our own—of rising corporate concentration and faltering competition; at such moments, Galbraith argued, markets could not be relied upon to police themselves. The solution he favored was a form of ecological balance: forces such as trade unions and consumer coalitions would act as a constraint. “Private economic power is held in check by the countervailing power of those who are subject to it,” Galbraith wrote. “The first begets the second.”
The past decade has seen the search for a countervailing power to offset the mighty tug of commercial interests. As Galbraith noted, government is not the only—or even the preferred—option; various other ideas have been mooted. In “Internet for the People” (Verso), Ben Tarnoff calls for a “deprivatized” Internet with new “models of public and cooperative ownership”; in “Own This!” (Verso), R. Trebor Scholz likewise explores the potential of worker-and-user-owned “platform co-ops.” (He discusses, although does not endorse, the idea of nationalizing large companies like Amazon and Facebook.) Dixon, in “Read Write Own,” reverts to a form of technological purism, resting his hopes in the potential of blockchain. The trouble is that, after more than a decade of casting about for checks on Big Tech, the only countervailing power seemingly able to muster the required heft and legitimacy is the nation-state.