The Age of Surveillance Capitalism

by Shoshana Zuboff

  To my Democratic colleagues and me, the digital tracks that a consumer leaves when using a network are the property of that consumer. They contain private information about personal preferences, health problems and financial matters. Our Republican colleagues on the commission argued the data should be available for the network to sell.147

  The reversal meant that although federal laws protected the privacy of a telephone call, the same information transmitted by internet immediately enters the ISPs’ surplus supply chains. This roust finally signaled the end of the myth of “free.” The Faustian pact that had been sold to the world’s internet users posed surveillance as the bitter price of free services such as Google’s Search and Facebook’s social network. This obfuscation is no longer tenable, as every consumer who pays his or her monthly telecom bill now also purchases the privilege of a remote and abstract but nevertheless rapacious digital strip search.148

  New and established companies from every sector—including retail, finance, fitness, insurance, automotive, travel, hospitality, health, and education—are joining the migratory path to surveillance revenues, lured by the magnetism of outsized growth, profit, and the promise of the lavish rewards that only the financial markets can confer. We will explore many examples drawn from these sectors in the coming chapters.

  In another trend, surveillance in the interest of behavioral surplus capture and sale has become a service in its own right. Such companies are often referred to as “software-as-a-service” or SaaS, but they are more accurately termed “surveillance as a service,” or “SVaaS.” For example, a new app-based approach to lending instantly establishes creditworthiness based on detailed mining of an individual’s smartphone and other online behaviors, including texts, e-mails, GPS coordinates, social media posts, Facebook profiles, retail transactions, and communication patterns.149 Data sources can include intimate details such as the frequency with which you charge your phone battery, the number of incoming messages you receive, if and when you return phone calls, how many contacts you have listed in your phone, how you fill out online forms, or how many miles you travel each day. These behavioral data yield nuanced patterns that predict the likelihood of loan default or repayment and thus enable continuous algorithmic development and refinement. Two economists who researched this approach discovered that these qualities of surplus produce a predictive model comparable to traditional credit scoring, observing that “the method quantifies rich aspects of behavior typically considered ‘soft’ information, making it legible to formal institutions.”150 “You’re able to get in and really understand the daily life of these customers,” explained the CEO of one lending company that analyzes 10,000 signals per customer.151

  Such methods were originally developed for markets in Africa to help the “unbanked”—people without established credit—to qualify for loans. One lending group interviewed potential customers in low-income countries and concluded that it would be easy to exploit the already beleaguered poor: “Most said they had no problem sharing personal details in exchange for much-needed funds.” But these app-based lending startups are typically developed and funded in Silicon Valley, and it is therefore not surprising that the same techniques have become part of a wider trend of exploiting American families that have been economically hollowed out by the financial crisis and neoliberalism’s austerity medicine. As the Wall Street Journal reports, new startups such as Affirm, LendUp, and ZestFinance “use data from sources such as social media, online behavior and data brokers to determine the creditworthiness of tens of thousands of U.S. consumers who don’t have access to loans,” more evidence that decision rights and the privacy they enable have become luxuries that too many people cannot afford.152

  Another example of surveillance-as-a-service is a firm that sells deep vetting of potential employees and tenants to employers and landlords. For instance, a prospective tenant receives a demand from her potential landlord that requires her to grant full access to all social media profiles. The service then “scrapes your site activity,” including entire conversation threads and private messages, runs it through natural language processing and other analytic software, and finally spits out a report that catalogues everything from your personality to your “financial stress level,” including exposing protected status information such as pregnancy and age. There is no opportunity for affected individuals to view or contest information. As in the case of digital lenders, although a prospective tenant must formally “opt in” to the service, it is those who have less money and fewer options who are trapped in this Faustian bargain in which privacy is forfeit to social participation. “People will give up their privacy to get something they want,” celebrates the CEO of this service firm.153

  Another genre of SVaaS firms employs data science and machine learning to scour the internet for behavioral surplus about individuals, either to sell it or to analyze and fabricate it into lucrative prediction products. Legal scholar Frank Pasquale describes this as “the dark market for personal data.”154 For example, hiQ markets its prediction products to corporate human resources professionals. It scrapes the web for information related to a client’s employees, including social media and public available data; then its “data science engine extracts strong signals from that noise that indicate someone may be a flight risk.” Machine learning models assign risk scores to each employee, enabling clients “to pinpoint with laser-like accuracy the employees that are highest risk.…” The company claims that it provides “a crystal ball” and that its predictions are “virtually identical” to observed turnover. With hiQ’s information, companies can preemptively intervene. They might make an effort to retain an employee, or they may choose to preemptively terminate someone who is predicted to be a “flight risk.”155

  Another example is Safegraph, a company that partners with all those apps that are tracking your behavior to amass “high precision/low false positive” data collected “in background from large populations.” According to the Washington Post, the company collected 17 trillion location markers from 10 million smartphones in November 2016 alone, data that were sold to two university researchers, among others, for a detailed study of political influences on patterns of family behavior on Thanksgiving Day that year.156 Despite the widely employed euphemisms of “anonymization” and “deidentification,” Safegraph tracks individual devices and the movement of their owners throughout the day, producing data that are sufficiently granular to be able to identify individuals’ home locations.

  Surveillance capitalism was born digital, but as we shall see in following chapters, it is no longer confined to born-digital companies. This logic for translating investment into revenue is highly adaptive and exceptionally lucrative as long as raw-material supplies are free and law is kept at bay. The rapid migration to surveillance revenues that is now underway recalls the late-twentieth-century shift from revenues derived from goods and services to revenues derived from mastering the speculative and shareholder-value-maximizing strategies of financial capitalism. Back then, every company was forced to obey the same commandments: shrink head count, offshore manufacturing and service facilities, reduce expenditures on product and service quality, diminish commitments to employees and consumers, and automate the customer interface, all radical cost-reduction strategies designed to support the firm’s share price, which was held hostage to an increasingly narrow and exclusionary view of the firm and its role in society.

  As competition for surveillance assets heats up, new laws of motion rise to salience. Eventually, these will shape an even-more-merciless imperative to predict future behavior with greater certainty and detail, forcing the whole project to break loose from the virtual world in favor of the one that we call “real.” In Part II we follow this migration to the real world, as competitive dynamics force the expansion of supply operations and an ever-more-complex extraction architecture reaches both further and deeper into new territories of human experience.

  Before we undertake that project,
it is time to stop and check our bearings. I have suggested that the dangers of surveillance capitalism cannot be fully grasped through either the lens of privacy or of monopoly. In Chapter 6 I offer a new way of thinking about danger. The threats we face are even more fundamental as surveillance capitalists take command of the essential questions that define knowledge, authority, and power in our time: Who knows? Who decides? Who decides who decides?

  CHAPTER SIX

  HIJACKED: THE DIVISION OF LEARNING IN SOCIETY

  They wondered why the fruit had been forbidden:

  It taught them nothing new. They hid their pride,

  But did not listen much when they were chidden:

  They knew exactly what to do outside.

  —W. H. AUDEN

  SONNETS FROM CHINA, I

  I. The Google Declarations

  On December 4, 1492, Columbus escaped the onshore winds that had prevented his departure from the island that we now call Cuba. Within a day he dropped anchor off the coast of a larger island known to its people as Quisqueya or Bohio, setting into motion what historians call the “conquest pattern.” It’s a design that unfolds in three phases: the invention of legalistic measures to provide the invasion with a gloss of justification, a declaration of territorial claims, and the founding of a town to legitimate and institutionalize the conquest.1 The sailors could not have imagined that their actions that day would write the first draft of a pattern whose muscle and genius would echo across space and time to a digital twenty-first century.

  On Bohio, Columbus finally found a thriving material culture worthy of his dreams and the appetites of the Spanish monarchs. He saw gold, elaborate stone and woodwork, “ceremonial spaces… stone-lined ball courts… stone collars, pendants, and stylized statues… richly carved wooden thrones… elaborate personal jewelry.…” Convinced that the island was “his best find so far, with the most promising environment and the most ingenious inhabitants,” he declared to Queen Isabella, “it only remains to establish a Spanish presence and order them to perform your will. For… they are yours to command and make them work, sow seed, and do whatever else is necessary, and build a town, and teach them to wear clothes and adopt our customs.”2

  According to the philosopher of language John Searle, a declaration is a particular way of speaking and acting that establishes facts out of thin air, creating a new reality where there was nothing. Here is how it works: sometimes we speak to simply describe the world—“you have brown eyes”—or to change it—“Shut the door.” A declaration combines both, asserting a new reality by describing the world as if a desired change were already true: “All humans are created equal.” “They are yours to command.” As Searle writes, “We make something the case by representing it as being the case.”3

  Not every declaration is a spoken statement. Sometimes we just describe, refer to, talk about, think about, or even act in relation to a situation in ways that “create a reality by representing that reality as created.” For example, let’s say the waiter brings my friend and me two identical bowls of soup, placing one bowl in front of each of us. Without saying anything, he has declared that the bowls are not the same: one bowl is my friend’s, and the other bowl is mine. We strengthen the facts of his declaration when I take soup only from “my” bowl and my friend takes soup from his. When “his” bowl is empty, my friend is still hungry, and he asks permission to take a spoonful of soup from the bowl in front of me, further establishing the fact that it is my bowl of soup. In this way declarations rise or fall on the strength of others’ acceptance of the new facts. As Searle concludes, “All of institutional reality, and therefore… all of human civilization is created by… declarations.”4

  Declarations are inherently invasive because they impose new facts on the social world while their declarers devise ways to get others to agree to those facts. Columbus’s declaration reflects this “conquest pattern,” as historian Matthew Restall writes:

  Sixteenth-century Spaniards consistently presented their deeds and those of their compatriots in terms that prematurely anticipated the completion of Conquest campaigns and imbued Conquest chronicles with an air of inevitability. The phrase “Spanish Conquest” and all it implies has come down through history because the Spaniards were so concerned to depict their endeavors as conquests and pacifications, as contracts fulfilled, as providential intention, as faits accomplis.5

  The Spanish conquerors and their monarchs were eager to justify their invasion as one way to induce agreement, especially among their European audience. They developed measures intended to impart “a legalistic veneer by citing and following approved precedents.”6 To this end the soldiers were tasked with reading the Monarchical Edict of 1513 known as the Requirimiento to indigenous villagers before attacking them.7 The edict declared that the authority of God, the pope, and the king was embodied in the conquistadors and then declared the native peoples as vassals subordinate to that authority: “You Cacics and Indians of this Continent.… We declare or be it known to you all, that there is but one God, one hope, and one King of Castile, who is Lord of these Countries; appear forth without delay, and take the oath of Allegiance to the Spanish King, as his Vassals.”8

  The edict went on to enumerate the sufferings that would befall the villagers if they failed to comply. In this world-shattering confrontation with the unprecedented, the native people were summoned, advised, and forewarned in a language they could not fathom to surrender without resistance in recognition of authorities they could not conceive. The exercise was so cynical and cruel that the approaching invaders often dispatched their obligation by mumbling the edict’s long paragraphs into their beards in the dead of night as they hid among the thick vegetation waiting to pounce: “Once the Europeans had discharged their duty to inform, the way was clear for pillage and enslavement.” The friar Bartolomé de las Casas, whose account bears witness to this history of Spanish atrocities, wrote that the Requirimiento promised the native people fair treatment upon surrender but also spelled out the consequences of defiance. Every act of indigenous resistance was framed as “revolt,” thereby legitimizing brutal “retaliation” that exceeded military norms, including grotesque torture, the burning of whole villages in the dark of night, and hanging women in public view: “I will do to you all the evil and damages that a lord may do to vassals who do not obey or receive him. And I solemnly declare that the deaths and damages received from such will be your fault and not that of His Majesty, nor mine, nor of the gentlemen who came with me.”9

  Conquest by declaration should sound familiar because the facts of surveillance capitalism have been carried into the world on the strength of six critical declarations pulled from thin air when Google first asserted them. That the facts they proclaimed have been allowed to stand is evident in the dispossession strategies of Verizon and other new entrants to the surveillance capitalist firmament. In the rapture of the young firm’s achievements, Google’s founders, fans, and adoring press passed over in silence the startling vision of invasion and conquest concealed in these assertions.10

  The six declarations laid the foundation for the wider project of surveillance capitalism and its original sin of dispossession. They must be defended at any cost because each declaration builds on the one before it. If one falls, they all fall:

  • We claim human experience as raw material free for the taking. On the basis of this claim, we can ignore considerations of individuals’ rights, interests, awareness, or comprehension.

  • On the basis of our claim, we assert the right to take an individual’s experience for translation into behavioral data.

  • Our right to take, based on our claim of free raw material, confers the right to own the behavioral data derived from human experience.

  • Our rights to take and to own confer the right to know what the data disclose.

  • Our rights to take, to own, and to know confer the right to decide how we use our knowledge.

  • Our rights to take, to own, to
know, and to decide confer our rights to the conditions that preserve our rights to take, to own, to know, and to decide.

  Thus, the age of surveillance capitalism was inaugurated with six declarations that defined it as an age of conquest. Surveillance capitalism succeeded by way of aggressive declaration, and its success stands as a powerful illustration of the invasive character of declarative words and deeds, which aim to conquer by imposing a new reality. These twenty-first-century invaders do not ask permission; they forge ahead, papering the scorched earth with faux-legitimation practices. Instead of cynically conveyed monarchical edicts, they offer cynically conveyed terms-of-service agreements whose stipulations are just as obscured and incomprehensible. They build their fortifications, fiercely defending their claimed territories, while gathering strength for the next incursion. Eventually, they build their towns in intricate ecosystems of commerce, politics, and culture that declare the legitimacy and inevitability of all that they have accomplished.

  Eric Schmidt asked for trust, but Google’s “declarations” ensured that it did not require our trust to succeed. Its declarative victories have been the means through which it amassed world-historic concentrations of knowledge and power. These are the bulwarks that enable its continued progress. Schmidt has occasionally revealed something like this point. When describing “modern technology platforms,” he writes that “almost nothing, short of a biological virus, can scale as quickly, efficiently, or aggressively as these technology platforms, and this makes the people who build, control, and use them powerful too.”11

  On the strength of its unprecedented concentrations of knowledge and power, surveillance capitalism achieves dominance over the division of learning in society—the axial principle of social order in an information civilization. This development is all the more dangerous because it is unprecedented. It cannot be reduced to known harms and therefore does not easily yield to known forms of combat. What is this new principle of social order, and how do surveillance capitalists take command of it? These are the questions that we pursue in the sections that follow. The answers help us reflect on what we have learned and prepare for what lies ahead.