Many ethical concerns related to computing and privacy are rooted in unstated beliefs about what human personhood is, in what context it applies, and what it entitles one to do. Why is it sometimes hard to talk in class, to state your ideas publicly, or to admit you don’t know something in front of your peers? Why is this easier (or harder) to do when it can be done anonymously? Why do social media platforms, whose stated main product is interconnection and whose source of revenue comes from that interconnection, give users the option to make their profiles private? Going further, why, in some cases, do these same sites give users the option to remove content including images or quotations that others have posted about them?

Most people in the West agree that a “right to privacy” exists, even if they don’t agree about what this right includes or how much it matters in practice. However, few people ask themselves why this right exists.

In fact, few people realize that the right to privacy was not conceptualized as something protectable by law until the end of the nineteenth century. Long before we ever posted selfies online, Samuel Warren and Louis Brandeis (Warren and Brandeis 1890) authored the first legal publication in the West to call for a “right to privacy.” This article has had an enduring influence on how privacy is imagined and understood. It was written primarily in response to the recent invention and proliferation of radical new technologies, including the simple, affordable box camera, the kinetoscope (motion-picture recording), and the phonograph (audio recording), which made it possible to document someone or something and to subsequently transmit, reproduce, and republish the documentation with or without the knowledge of the person being recorded.

Despite its foundational status in privacy debates, this legal opinion is a product of its era, written in response to the technological changes that were happening at a specific moment in time. Therefore, it does not always line up clearly, or indeed at all, with the potential incursions on privacy that more recent technologies have made possible. Similarly, Warren and Brandeis’s underlying assumptions about personhood are a product of that time. But because personhood is so rarely recognized as a core element of how we understand privacy, it can be harder for us to notice how these authors’ presuppositions surrounding personhood have carried over into contemporary debates. This inherited concept requires some re-examination, just as privacy does.

However privacy is defined, the implicit role of personhood is visible once you think to look for it. Generally speaking, the language of “privacy” is used to refer to an expectation that a person’s actions, words, and information about their identity will not be shared in certain contexts or with certain people or institutions without that person’s consent. Indeed, as technology and society scholar Shoshana Zuboff notes, Justice William Douglas stated in 1967 that “the individual should have the freedom to select for himself the time and circumstances when he will share his secrets with others and decide the extent of that sharing.” Zuboff sees this “freedom to select” as central to the epistemic right to know ourselves and how we are shared (Zuboff 2021). Even this very basic definition of privacy links it to self-determination: that is, a person’s ability to exercise some level of control over themselves, how they represent themselves to others, and how they choose to live out their life. Without an operational understanding of personhood, arguments for privacy would make little sense. But only rarely are these presuppositions stated and defended.

By grouping privacy and personhood together, this chapter underscores ways in which developments in technology change not only our understanding of what privacy is but also the ways in which technological developments more broadly speaking can alter our sense of what should be permissible for a person. But first let’s tackle the concept of personhood.

In our analysis in the previous section, it was suggested that when someone asks the question “What is a person? ” they may be asking about what counts as human life. The simple explanation for this is that for various cultures, past and present, “personhood” is a term that is used synonymous with “human being.” But this explanation may not please everyone—and there are good reasons to be cautious here.

If we automatically assume that the concept “person” can be used interchangeably with “human being,” then when we ask the question “What is a person? ” we may be asking about what is the most fundamental description of human life that we can think of. A common and indeed ancient way of answering this question has been to find a trait or property that humans are said to possess but that other living or nonliving entities do not. For instance, one might suggest that human beings have self-consciousness or that they use language, and that other living beings do not have these things; for this reason, what it means to be a person (i.e., a human) is to be a being that uses language or possesses the capacity for self-consciousness—that is, there is a quality intrinsic to all persons that all other beings lack. Although this might sound familiar or seem like a straightforward way to proceed, it tends to blur together metaphysical beliefs, epistemological presuppositions, and axiological commitments. This means it is hard to gain insight into what kind of answer we are expecting. Complicating matters further is the fact that not everyone agrees about what set of properties, characteristics, or traits are the most essential or fundamental to human life (this is an axiological question).

In the context of many Western societies, personhood is said to be what accords human individuals universal worth and inalienable rights. The notion that human persons have intrinsic rights historically has been used to define the rights of citizens against others and in relation to the state (Mill [1869] 1991; Rawls 1972; Rousseau [1762] 1997). Within this intellectual tradition, rights and duties are understood to be rooted in the human being’s status as, for instance, a self-conscious and rational agent. But historically speaking, not everyone was said to fully possess these capacities and traits. In some cases, these capacities were presumed absent from those of a particular gender, race, or class. For example, Aristotle, who is well known for his arguments regarding the rational status of human beings, denied rationality to women and claimed that Africans were naturally slavish due to living in warmer climates. This form of argumentation, known as “climate theory,” was used to justify and naturalize slavery for centuries (Kendi 2016).

These same biases shaped how personhood was accorded in the United States as well. The politicians and philosophers who molded the United States in its early decades were almost all of European descent. These men, who framed both the laws and norms of this new society, denied legal standing and hence personhood to Native Americans on the grounds that they were of a different nature or had not yet progressed enough developmentally to be fully human. Black Americans were denied equal standing because they had been declared property rather than persons (Kendi 2016). Hence, what appears to be a universal claim about all human beings has, historically, worked in ways that exclude certain people and include only a select few, and the practical implications for those excluded are broadly, profoundly, and enduringly harmful.

Despite this unsettling legacy, intrinsic notions of personhood are quite prevalent even today and are frequently called upon to support human dignity and to argue for the value of human life in all its forms. Indeed, Article 1 of the Universal Declaration of Independence, for instance, states that “All human beings are born free and equal in dignity and rights. They are endowed with reason and conscience and should act towards one another in a spirit of brotherhood” (United Nations 1948). Article 6 further specifies that “Everyone has the right to life, liberty and security of person” (United Nations 1948).

Intrinsic notions of personhood are not the only way of conceiving of personhood, however. For instance, political philosopher Martha Nussbaum suggests that human dignity and the respect that we owe to others, at least from a political perspective, need not be based on controversial epistemological or metaphysical assumptions but rather on political commitments and, with time, overlapping consensus (Nussbaum 2011; Formosa and Mackenzie 2014).

Furthermore, as we saw in chapter 2, sub-Saharan African communities largely understand personhood as a community matter: personhood is not intrinsic but is rather something achieved in and by means of a given community (Menkiti 1984). Furthermore, these traditions generally do not ascribe major metaphysical significance to humanity, as distinct from plants, animals, natural phenomena or spirits. In the context of this “cosmic humility” (Tangwa 2005), personhood is a social concern that is of practical importance to other persons who live in the same community but is not necessarily a matter of unique philosophical significance.

Leaving aside the presupposition that personhood is fundamentally a question about what it is that defines human identity or a question about what qualities or characteristics are most fundamental for human life, we can approach the topic of personhood from a different angle.

You might feel that the best way to attack a big, nebulous concept like personhood is to “solve” for a smaller part of it; in other words, figure out what makes an individual person distinctive and then generalize from that to persons in general. But of course, we aren’t just smaller units of personhood; each of us is made distinctive by our personal identity.

Your “identity” can be broadly defined as certain demographic facts about you that you may have little to no control over—for instance, that you are a human being, the location of your birth, your ethnicity, and the gender that you were assigned at birth. It may also refer to things over which you (in most cases) have some control: your credit score, your marital status, and your religious affiliation.

Personal identity, however, points to something more specific—something, well, more “personal.” Personal identity is typically understood as those aspects, characteristics, and features that define a person’s existence in relation to all others—the things that make you who you are. What makes you who you are can involve things that you have chosen for yourself, such as your actions and beliefs, as well as other things that you may not have chosen that influence you.

For instance, we may consider our relationships with others, our place within society, and/or the communities in which we live to be part of our identity. It is also common to think of personal identity as having something to do with the choices that we have made, the way we live our lives, and the way we define ourselves in relation to others.

Your identity may also include certain details about your temperament, your aesthetic preferences and other psychological traits. As we saw in our discussions of both virtue ethics and communitarianism, a person’s present self is largely a product of influences that they have not necessarily chosen or that affect them in ways beyond their control: where they grew up, the experiences they have had, and the people with whom they spend their time. Aesthetic preferences—that is, questions of personal taste—are an interesting example of how each of us is shaped by our time and place. It would be ridiculous to claim that our tastes are only a matter of when and where we come from: after all, it’s unlikely that you and your friends all have identical taste in music or movies or food, among the various options on offer in the present. You may even know someone (or be someone) who particularly enjoys mid-twentieth-century noir films, or who refuses to listen to music made after the early 1990s. All the same, our preferences are shaped, at least in part, by what we are used to and what is available to us.

But if we focus only on these things—the things that make you who you are or make me who I am—the question of personhood becomes primarily about self-knowledge (who am I?). But surely this cannot be the only way to think about personhood because, as stated above, personhood is, in most cases, a normative concept.

If personhood is the “quality or condition of being an individual person,” clearly what it means to be a person or the quality or condition of being a person is not rooted only in the way that you define yourself but is rather a larger concept that we can apply to others. In other words, we cannot use the same things that you uniquely attribute to yourself to define what personhood means more generally. Personhood must encompass both the possibility of those individual differences and some underlying commonality.

It was noted above that personhood is sometimes understood to be an intrinsic property—a property that all human beings share. Let’s assume for a moment that each of us has a unique identity or essence that makes us who we are. Is this unique identity something that we achieve over time? Is it something that can be lost? Is it something that can be duplicated or repeated? Or is it possible only for one individual to have one unique identity? In the Western philosophical tradition, these questions take us to the very heart of the matter of personal identity.

Historically, the debates about personal identity in this Western context are centered around the following two main themes:

• A search for a principle of individuation, that is, a principle or rule by which something can be distinguished from other things or from others within the same group or classification.
• A way to account for the continuity of identity, that is, what allows something to acquire and maintain an identity over time.

Both of these themes might seem far removed from ideas about personhood, but not if we consider some of the questions that come up in popular discourse about technology. For instance, is it possible to upload our minds into machines? Do clones have the same identity as their originals, and if not, how do they differ and why? Can we replace most of our organs and limbs with technological prosthetics and still have the same identity?

Although these specific questions may reflect the contemporary popular imagination, philosophers have long wondered about what constitutes personal identity and how it can be lost, achieved, or maintained. We will touch on some of these ideas so that in the next section we can see how ideas about human identity and personal identity shape the ways we think about technological change and mediation.

Theories of individuation look for the cause or principle that can account for the distinct nature of an individual entity. A principle of individuation can be used to explain how humans can be distinguished from other things (chairs, dinosaurs, rainforests, microwaves); alternatively, it can be used to explain how persons are distinguished from other persons. The principle in question is not something that pertains to specific traits like the type of haircut you have, or the kind of music you like, or where you grew up. It is said to be a fundamental or essential principle, cause, or mechanism that makes it possible for there to be a fundamental distinction between individuals. It amounts to a kind of metaphysical claim about what something is in its essence and how that essence can be distinguished after it has been assigned a category, such as human beings or rational agents (Gracia 1984).

For many ancient thinkers, and still for many people today, what makes a person a person hinges on a metaphysical belief about an intrinsic property that occurs only within human persons, for instance, a soul or an essence.

In the context of the Western intellectual tradition, historically speaking, the principle of individuation was about a substance of some kind, something either material (for instance, the body) or immaterial (for instance, the soul or intelligence) that could endure through time. This enduring substance accounts for individual personal identity, because the individual is one and the same substance over time. Whether that consistent substance is material (like a body) or immaterial (like a soul), the continuity of the substance helps to explain the continuity of the distinct individual composed of it.

When this view is maintained, there is no real need to determine what remains continuous over time since what remains continuous, as well as what differentiates one being from other beings, comes down to an essential substance. But not all theories of personal identity look for a principle of individuation to determine what constitutes personal identity. And not all theories of personal identity rely on metaphysical claims about an underlying substance or essence.

In fact, there have been various shifts in the way personal identity has been understood within the context of philosophical, legal, and political thought. In sixteenth-century Europe, the term “person” was frequently used to talk about participating in a community and being subject to that community’s laws. (You may notice a resemblance here to how “being a person” is understood in sub-Saharan communitarian traditions.) Thomas Hobbes, for instance, tied the concept of person to representation and how we appear to others. He appealed to the Latin roots of the term: “persona” in Latin originally referred to the theatrical mask used in both Greek and Roman theater. Accordingly, Hobbes was able to describe two different kinds of persons: natural persons who represent themselves and artificial persons who represent other persons or things (Hobbes [1668] 1992).

David Hume, notably, argued that personal identity doesn’t have any enduring status at all (Hume 1978). For him, if we think there is an underlying substance or unity that supports the continuity of the self, we are mistaken. The mind, he insisted, is nothing more than a bundle or collection of different perceptions. Individuals are differentiated by their perceptions and experiences but there is no essential self onto which these perceptions and experiences adhere. According to Hume, although we have a tendency to think of ourselves as the same person that more or less remains the same over time, this is nothing more than an illusion—a trick of the mind.

John Locke (whom you may remember from our discussion of utilitarianism) had similar ideas about personal identity. However, his account is particularly noteworthy because he is often cited as the one who cut the knot that for centuries tied personal identity to metaphysical ideas about substance. For him, the question of personal identity is not a matter of determining “what the person is made of” but rather “what it is that allows a person to maintain their identity over time.” He insisted that the identity of a person is not to be based on a material substance such as the human body nor does it rely on an immaterial substance such as a soul or spirit. Instead, the identity of persons rests solely on consciousness. A person, Locke maintained, is best defined as “a thinking intelligent being, that has reason and reflection, and can consider itself as itself, the same thinking thing in different times and places; which it does only by that consciousness” (Locke [1690] 1990, II, §26).

Locke also thought it was important not to conflate our understanding of persons with our formal definitions of humanity (Thiel 2011). To refer to a person is to refer to an intelligent agent that is able to extend “itself beyond present existence to its past, whereby it becomes concerned and accountable” for past actions (Locke [1690] 1990, II, §26). The unity and identity of the person is a continuous process through which the past is brought into the present.

Locke understood the idea of persons to be a “forensic” notion, which is to say that he thought that our ideas about personhood were crucial for understanding how we ascribe moral and legal responsibility. In order to ascribe actions to an individual, we have to be able to see them as the cause of actions. In order to ascribe moral and legal responsibility to a person, that person needs to be seen as a coherent and consistent self. In Locke’s judgment, without memory and without the capacity to reflect on our own actions, this type of consistency and coherency is impossible. In other words, Locke’s approach to personal identity was unique in part because of the emphasis it placed on self-awareness or self-consciousness.

It is worth noting that self-consciousness is another of those traits or capacities that are often used as a way to distinguish persons from other entities or things. Not surprisingly, this also means that people continue to debate its exact meaning (Crone et al. 2012).

Generally speaking, self-consciousness refers to a capacity that makes it possible for individuals to distinguish themselves from the world in which they are embedded. Self-consciousness is also what makes it possible for individuals to recognize feelings, thoughts, and emotions as their own. Some philosophers have argued that self-knowledge—knowledge that we gain about ourselves—presupposes self-consciousness and, therefore, it is an essential aspect of what it means to be a person. Others argue that self-consciousness is necessary for us to deliberate about our actions and intentions and, therefore, it plays a fundamental role in ethical thinking (Korsgaard 1996).

As noted above, Locke’s conception of the human person is forensic; that is, it is concerned with how it is that we attribute responsibility. This is one of the reasons that self-consciousness is so important for Locke’s conception of persons. Without it, a person could not be understood to be the author or agent of their actions and therefore could not be held accountable for their actions. As influential as this approach to personal identity may be, it prioritizes the human’s capacity to know and do.

A different perspective on personal identity can be found in what is sometimes described as the narrative approach to personal identity (McLean 2008). Narrative identity begins with the basic claim that human beings are not simply knowers and doers—they are also storytellers. As both highly intelligent and imaginative creatures, humans use narrative to explain where they come from, where they are, and who they wish to become. These narratives reflect aspects of one’s life that they find meaning and valuable. Because narrative plays such an important role in human life, theorists across various fields have used the term “narrative identity” to capture something about what it means to be a person that other accounts leave out. As DeGrazia explains,

Your narrative identity involves your self-conception, your self-told story about your own life and what’s most important to you. Rather than merely listing events involving you or facts about you, it orders these events and facts, highlights certain features and people, and organizes what’s highlighted into a more or less coherent story of your life and character. (DeGrazia 2005, 179)

The experiences, events, and happenstances of our existence can be construed as information about who we are as people. But they do not seem to present us with a complete picture of who we understand ourselves to be. Narrative identity helps capture the fact that our identities are not something that happen to us: they are something that we creatively express as we talk about ourselves and share the events of our lives with others. Self-knowledge, in other words, is not something we can gain through immediate and direct analysis. Instead, narrative identity theorists argue that self-knowledge is something we gain through mediation. In particular, narrative identity theorists argue that we gain self-knowledge through the stories that we have been told, or through the symbols and myths that we have inherited through different social settings (Ricoeur 1991). Therefore, it is often the case that our feelings and judgments about how best to represent ourselves reflect the values and prevailing story lines of our own sociocultural context.

In most classically liberal and democratic nations, individual freedom and individual control over one’s actions, body, and choices are considered valuable and worth protecting. Accordingly, these societies typically recognize certain rights in order to protect them. With this individual freedom and control comes responsibility, both legal and moral. Individuals are understood to be the authors of their own actions and therefore are responsible for those actions.

What does responsibility entail? Although moral responsibility and legal responsibility are closely related, they do not always carry the same meaning. In this section we will identify how they overlap and why in some instances they differ.

Legal responsibility is assessed (and intentionally so) according to criteria that closely align with moral responsibility. One such criterion is that the individual in question, in order to be legally culpable, must have moral agency in order to be responsible. Generally speaking, moral agency is comprised of the following five aspects:

• Agency: Defined as the capacity to act and effect change on our environment and to affect other living and nonliving entities.
• Self-determination (autonomy): The idea that we can act in ways that are not fully determined by the environment in which we find ourselves nor by our historical or social context.
• Self-awareness: An awareness of one’s self in distinction from objects and from other agents acting in the world.
• Rationality: Narrowly defined as the cognitive process through which we determine that for every action or event there is a cause.
• Responsibility: The capacity to identify our actions as our own and take responsibility for those actions, and/or the capacity to accept that we are part of relations through which we come to determine what we are responsible for and what we are not responsible for.

Another important criterion for legal responsibility, which likewise overlaps with many ethical frameworks, rests on the distinction between voluntary and involuntary actions. As we discussed in chapter 2, in the context of most traditional Western ethical frameworks such as deontology, virtue ethics, and utilitarianism, moral responsibility and culpability apply only in those cases in which voluntary action is performed. Involuntary actions (understood in this context as actions performed out of ignorance or action compelled by force or coercion [Aristotle 1999, III.1]) may be considered good or bad with respect to their consequences, but the moral agent is responsible only indirectly, if at all, for involuntary actions (Sidgwick [1874] 1981). Similarly, a moral agent is less responsible—if at all—for involuntary action (which typically means action performed out of ignorance or action compelled by force or coercion [Aristotle 1999, III.1]) than for voluntary ones. Similarly, most criminal offenses require proof not only of an actus reus (a “guilty act”) but also of a mens rea (“guilty mind”), a specific state of mind on the part of the accused. Furthermore, an actus reus must include a voluntary act for someone to be charged with a crime. As the legal theorist and philosopher Anthony Duff explains: “In the absence of a voluntary act, any movements of my body are involuntary; and whilst those movements might cause harm, thus making me (or them) causally responsible for the occurrence of that harm, I lack control over those movements and cannot justly be held either morally or legally responsible for their effects” (Duff 2009).

Yet although the criteria overlap significantly, it’s not hard to find situations in which legal and moral responsibility do not align. For example, it is not illegal to entice children to spend money on virtual goods or hook people into pay-for-play platforms, but there may still be moral reasons for finding ways to prevent these practices (Hohn 2018).

On a more personal level, most of us can recall times when we felt morally guilty or responsible for something we did even though we did not break the law. Even if we cannot recall a specific time we felt this way, we can likely think of examples in which people explained their actions in terms of guilt and remorse. Similarly, most people would agree that while, for example, jaywalking and stealing money from your neighbors are both illegal actions, they do not necessarily come with the same moral weight. The point is that even if we believe in some cases that the parameters of legal responsibility could be fine tuned or updated to match our own moral judgments, the existence of these gaps helps to underscore that they are not the same thing. Legal theorists, philosophers, sociologists, and psychologists have offered various explanations for how and why they are different. One fairly common suggestion is that moral responsibility can be distinguished from legal responsibility because it is more deeply concerned with an agent’s attitudes and responses. When we hold a person (whether ourselves or another) to be morally responsible, we evaluate them not only on the basis of the things that they do or don’t do, but also on the basis of the underlying dispositions that shape the actions and attitudes that they express about others’ actions (Strawson 1962; Watson 2004). Blaming somebody in a moral sense (as opposed to a legal one) involves passing judgment on their character. It involves finding fault with their ideas about good or bad ways to act, or their responses to external actions or events (Nozick 1981). These shortcomings or defects do not imply illegal or criminal behavior, but they may still be considered failures of moral responsibility. Imagine, for example, that you learn that a friend of yours has died in a horrible accident. When you call to inform one of your mutual friends and share the date of the funeral, he acts indifferent to the news and says he can’t attend the funeral because he can’t miss soccer practice. Although this might be a simple matter of bad judgment, it is an example of a situation in which we expect certain responses from people, not only with respect to how they act but also how they feel about certain things. We can be let down or change our opinion about their character when they do not respond in the ways that we have come to expect.

Another difference is that the consequences of legal responsibility are also more concrete: the legal system holds you accountable, or doesn’t, as the case may be. The consequences of moral responsibility are far less straightforward. For instance, it may briefly feel rewarding to tweet “history will judge you harshly” at somebody whose actions you condemn, it’s difficult—perhaps impossible—to enforce a felt sense of moral responsibility in others. The simple reason for this is that you cannot control how people will respond to certain situations, let alone control how they feel, at least as an individual. This sort of enforcement does often take place at a broader social level: communities or societies always express expectations about how people should think and act. These expressed expectations can be very powerful, but there are no clear ways to measure or fully predict their impact on a particular human individual.

Various works of science fiction invite us to consider the personhood of androids so advanced and human-like that they can pass unmarked in human society; these stories seem to urge us, by and large, to expand our definition of personhood to include these not-quite-humans or warn us of the dangers of technology so advanced as to be worthy of personhood. While it may be true that such advanced technology is many horizons ahead of us, these stories also push us to consider the scope and boundaries of personhood.

Humans have been grappling with questions concerning the personhood of nonhuman entities, such as advanced artificial intelligence (AI), for some time now. In fact, societies have considered the personhood of animals and other nonhuman entities for centuries! The notion of building an autonomous entity goes back very far: many ancient Greek myths imagine the construction of autonomous beings, including but not limited to humans themselves. The Jewish myth of the Golem, an anthropomorphic creature built out of mud, goes back at least to antiquity. After the fall of Rome, widely circulated tales reported that the Empire had left behind vast underground troves of wealth, guarded by golden automata. But only in the nineteenth century did the stories of such built entities start to consider the potential personhood of these entities and the ethical problems that their personhood would entail. Perhaps one of the best-known stories is Mary Shelley’s 1818 novel Frankenstein; or, The Modern Prometheus, which is really a story about the creature (the Monster) as a person rather than the scary-monster story of later film adaptations. The word “robot” was introduced by Czech writer Karel Čapek in a 1920 play, R.U.R. The play includes notions of the “mechanization” of labor (though Čapek’s “robots” are more biological than mechanical) and includes ideas that are now referred to as the “robopocalypse,” wherein robots destroy humanity.

It can be very tempting to jump right into debating whether robots should count as people. Many mainstream news stories and think pieces about advanced AI systems are organized around this exact question. Before we start down that road, consider other, more familiar nonhuman entities that could (and perhaps should) provoke the same question. Crows, for instance, make tools out of sticks and their own feathers and drop stones into containers of water to raise the water level. Elephants take their young to visit the bones of their ancestors. Elephants and crows are not human beings, but they use tools and engage in burial practices, both of which have been suggested as markers of a distinctly human life (Jonas 1996).

One of the reasons it is both strange and familiar to think of nonhuman entities as having personhood is that there is a long tradition in which nonhuman entities are said to have some kind of legal or political standing. In ancient cultures animals were charged with and punished for crimes. In the United States, corporations are considered to be legal persons. Legal personhood often determines what sort of entities are deserving of certain rights and protections. Should certain forms of artificial life be granted with legal personhood? In 2017 the robot Sophia was granted citizenship in Saudi Arabia (Reynolds 2018). Earlier that year, the European Parliament adopted a proposal to explore the implications of how robots are classified legally and the possibility of creating a special legal status for them. The possible consequences of applying an existing framework like corporate personhood to robots are complicated: for example, the possibilities that such a law could allow humans to shield themselves from liability for robots’ violations of human legal rights (Pagallo 2018). (See the sidebar “Metaphors, Personhood, Technology, and Autonomous Cars” in chapter 6 for some of the complications with defining what should even be regulated as a “robot” to begin with.)

Even if we don’t think that technological objects of any kind can have personhood, there is still the matter of anthropomorphism—of attributing what we identify as human-like qualities to nonhuman entities such as a boat or a robot—and the effects of this practice. Darling (2017) details an experiment with two groups: one group that interacted with a robot who was given a human name and a backstory, and another group that was given just the robot with no backstory. The group given the backstory showed more standard markers for empathy when interacting with their robot, including asking if it was okay when struck with a falling object. This study suggests that the framing of the robot relationship can influence how people perceive, and subsequently interact with, a given piece of technology.

There are arguments against encouraging anthropomorphizing technology, especially forms of social robots that are specifically designed to interact with humans such as caregivers or teachers. The predominant argument is that these technologies may “replace” relationships with real humans because interacting with robots may be “easier” in some cases (Turkle 2010). For instance, roboticist Carme Torras, in her novel Vestigial Heart (Torras 2018), imagines a world in which each person has a helper robot rather than meaningful personal relationships. However, as Darling (2017) points out, there are many examples in which instead of replacing a relationship with a human, a robot can augment and enhance a human relationship. Consider an instance of an authority figure and a robot: Darling details specific cases including using both a Paro, baby seal robot, and a social worker in which interviews with young children were more productive with both the human and the robot present. Hence, social robotics can sometimes encourage and facilitate communication.

Personal assistants have become more commonplace in the home. Many people have started to think of the disembodied voice of Alexa or Siri as part of their households, another person with whom to discuss, interact, and ask to play music. These digital assistants are oftentimes deployed within a home with multiple users, be they parents and children or a set of roommates. This situation of an entity within a shared space raises a number of curious questions related to the personhood, and more specifically the responsibility, of digital assistants. Consider a case in which two younger children are home alone and a friend brings beer to the house for all of them to drink: should the digital assistant immediately call the parents? Does the assistant have a responsibility to ensure the safety of the children? Or is it subservient to their desire to drink beer and have a good time breaking the rules? Or consider the simple question of a shared digital assistant between two partners: if the agent is instructed by one partner to lie to the other about a date or a location, should it? Can an assistant read your partner’s email to you? These questions are explored in detail by Luria et al. (2020), who find that many of these questions largely come down to how we perceive the social role of the technological artifact itself. Do we see the assistant as a full person, a member of a family? Is the assistant “owned” by a particular person? Or does the assistant have responsibility to everyone within the home? Does the way in which you define the personhood of the assistant make a difference in deciding to whom or to what the assistant owes responsibility?

These examples show us that it is important to be aware of how the language we use, or are encouraged to use, to refer to technological artifacts may influence both our interactions with and our thinking about these artifacts and their personhood.

As we have seen, some definitions of personhood rely on metaphysical beliefs about an underlying substance that is said to constitute the unity and continuity to individual entities. We have also discussed how memory and self-consciousness can function to support notions of personal identity. Perhaps you have heard the hypothetical situation in which humans will eventually be able to upload their consciousness or their memories to a computer and “live” forever. How do the definitions of personhood provided above operate in ways that support these kinds of claims? How do they limit them?

Consider how our definitions play out in an example in which one consciousness, or at least appearance of consciousness, has the ability to “jump” between bodies. You may have had an experience close to this already if you’ve interacted with Alexa, embodied in multiple different speakers throughout your home, or Siri on your iPhone versus your laptop. Consider now a computer assistant/persona that moves between locations and even embodiments while interacting with you. There are multiple ways that this assistant can inhabit various embodied forms, which is totally different from what we as humans can do. Consider the following options for the embodiment of an assistant:

• One-for-one: Each assistant has a single physical embodiment. This is closest to our experience as humans.
• One-for-all: One assistant is embodied in several forms at the same time. This would be like talking to Siri on your smartphone and then interacting with Siri on your laptop. The physical embodiment and modes of interaction may change but the “consciousness” inhabits all of these embodiments simultaneously.
• Re-embodiment: One assistant is embodied in several forms in sequence, but only one form at a time. This is similar to the Siri example but instead of being able to control multiple embodiments at the same time, Siri would have to “move” between them.
• Co-embodiment: Multiple assistants can be embodied in the same physical form. This would be akin to interacting with both Siri and Alexa on your smartphone, at the same time. Both consciousnesses inhabit the same physical embodiment.

How do these options change your perception about the personhood of the digital assistant? Would you treat all these the same? Would the embodiment change how you perceived the assistant or what it could do? One such experiment was carried out first with a robotic butler that helps to check you into your room at a hotel reception, in the form of a humanoid robot with a name and identity that connects to you. The persona then could jump to different artifacts later, a TV screen in another room or a speaker overhead, to help you as you moved through a new space. How would you feel about interacting with such an entitity? It might seem nice to have one “personality” follow us between locations, but it also seems that it could be unsettling in some contexts (Luria et al. 2013; Reig et al. 2020).

This technology in particular presents a challenge for some views of personhood. Specifically, it directly puts in tension the distinction between the material substance-based metaphysics (i.e., that people have the same body over time), and the nonsubstance-based view (i.e., that people have memory and self-awareness). In the case of the body-hopping assistant, it turns out that individuals’ reactions depended heavily on the context. In the example noted above involving a robotic butler, people reported enjoying the fact that the personal assistant could follow them around and look after them in such a personalized way. However, when the same experiment was performed at a hospital, where the robotic concierge checked them in and then jumped to a smart TV in the room to offer assistance on dosing for medications, many participants thought that the assistant was not “smart enough” to give advice on dosing because it was working the front desk/check-in, saying “when you go to the X-ray room, [you want to] have a professional person [social presence] there, not [the one from] the reception area” (Luria et al. 2013).

As we have noted above, one’s conception of one’s self is shaped by our life experiences and the particular ways in which we have internalized these experiences. This requires consciousness, certainly, but also the capacity to retain and remember past events. Insofar as your memories play a crucial role in shaping your goals and your values, they are an important part of your ability to think and act in a way that feels like it is “really” you.

It can be tempting to think that this “real” you, the one that is informed by your memories, somehow exists apart from the material world, including technology. But as we have noted, the boundary between the self and the world is not so easy to pin down in practice. Vision-correction technology or hearing aids can seem external to the self, but they can have a huge impact on how a person perceives and engages with the world. Similarly, technologies designed to collect, store, and transmit memory in many ways determine the means and conditions under which a person interacts with their own memories, either by storing things externally or by “remembering” our passwords and preferences so that we ourselves do not have to reproduce them regularly.

The outsourcing of “remembering” to smart technologies might seem like a threat to “real” you, but trying to remember a time when human beings did not rely on external memory supports seems impossible (as we discussed in chapter 3). In many ways, memory is experienced and realized differently depending on the technologies that are available and considered “normal.”

To put this into perspective, consider Locke’s notion of personal identity above, which emphasizes the importance of continuity of the self by reference to consciousness and memory. If memory is taken to be one of the fundamental aspects of our definition of personhood, is there more at stake when we rely on external memory supports? If, for instance, one’s personal memories and experiences could be transferred into a different entity, does that person exist in two places? Do they really die when their body dies? As more and more of our memories are stored in images, text, and other digital forms, some people have wondered: do multiple copies of ourselves exist? Answering these questions requires us to gain clarity about our own ideas about what constitutes personhood.

Up to this point, we have assumed a “disembodied” online experience, in which other people perceive a user only through content that user has written or uploaded. But since physical appearance is such a powerful component of personal identity for many people, it is not surprising that developers have tried to recreate the experience of embodiment virtually as well.

In the three examples provided in the previous section, privacy is understood as a separate sphere, space, or place for reflection. In the first example, workers who are continually exposed to others in an open environment are unable to perform their work and maintain a healthy sense of well-being. In the second example, gender identity moves from the private or personal sphere to the public sphere when someone rejects or claims a particular gender identity. In the final example, incarcerated people are denied the right to a physical space where they are free from surveillance and coercion. Each of these examples highlights how removing access to privacy as a discrete space fundamentally alters someone’s sense of personhood, in their own eyes and in others’.

For most of recorded history, privacy has been conceptualized in terms of physical space. Aristotle, for instance, made a distinction between public and private spheres (Shields 2022; DeCew 2018). He made it clear that there are some things that are meant to be shared broadly, and some things that are not. This distinction between private and public was also important during the seventeenth and eighteenth centuries in the context of the European Enlightenment. In the Victorian era (a designation for British history and culture during the rule of Queen Victoria, 1837–1901) the financially well-off classes, professional or rich, distinguished between the public and private spheres (Digby 1993). Much as for the ancient Greeks, for the Victorians the public sphere encompassed commercial, political, and nonfamily social interactions (Arendt 1958). According to this classification of space and social function, the public sphere was exclusive to adult males (Digby 1993). The private sphere was physically located within the house and included the management of the household and of relationships within the household. This of course implies that anything that happened within the context of the household was “private.” This did not mean that servants or family members were confined to themselves. It meant that the spaces were “private” because certain people were excluded from that private space and those “private” interactions. Likewise, those people who were seen as primarily belonging in the private sphere—women, children, servants, and slaves—were rarely, if ever, able to participate in public life. Indeed, in Greek culture, nothing was more embarrassing to a public figure than to have their private life, such as their wife or their family, discussed in public spaces (Garland 2009).

The literal division of spaces helps us separate some aspects of our physical, daily lives: the home, the office, or public institutions. And as noted in sections 4.3.7 and 4.3.8, because we are not physically present when we interact with others online, we can conceal parts of our identity that would be difficult to conceal in other types of interaction. But the organization of the world into spheres or spaces was not and is not only a literal configuration: it also functions conceptually. Although this distinction doesn’t specify the legality of what information can be obtained, it does enable us to think about what it means to be private by helping us think about whom and what “privacy” excludes.

When we talk about the division of online “spaces,” we are talking about a conceptual division, using physical space as an analogy. It can help us think about the way we perceive the space of our interaction, who is included, and who is not. You can see the space analogy at work even in dealing with the operating system of your computer or the design of networks. Access is partitioned into “zones”; networks exposed to the outside are called “demilitarized zones,” or DMZs, to denote when the network has gone from a private local intranet to the public internet (Kurose and Ross 2016; Silberschatz et al. 2008).

However, this analogy of space can break down in important ways that we often do not realize, and this breakdown can cause us to misinterpret the degree and kinds of privacy that are actually available in the places we access virtually (Betz and Stevens 2013). For instance, email servers, both online like Gmail or available through your school/workplace, may seem like private spaces: they have all the trappings of a private space, including locks, such as login screens, to restrict access. But in general, these spaces are not exclusively yours. In the United States or the European Union (EU), if you work for a government or a private company, you have no claim to privacy over the email that you write or send through the employer’s servers (Privacy Rights Clearinghouse 2019). Similarly, for most online providers, algorithms read your email to place ads and direct you to relevant content on the internet. Likewise, with the proliferation of digital assistants in the late 2010s, such as Amazon Alexa and Google Assistant, the words spoken by you in a number of spaces are recorded and transmitted to servers for processing.

As we have seen, even in ancient times there was an idea of a distinction between our bodies, our homes, and property versus things that belong to the social, the public, or the state. But over time privacy has also expanded to include a sense of control over information—information about oneself. It is not all that surprising that privacy is associated with control over information. Notice that in all three of the examples given in section 4.5.1, control of various kinds is essential for privacy: control over how one presents oneself socially, control over one’s interactions with others, and control over how one organizes one’s time. This focus on control in many ways echoes the division of public and private: things that are private are the ones that we “own” or “are master of,” in contrast to those that are public.

This sense (real or perceived) of ownership does not reflect only our practical control; it is quite common for people to see one’s possessions as either part of or reflective of one’s sense of identity. The car we drive, the clothes we wear, the kinds of books and art objects we own, the equipment that we use to play certain sports, or the specific tools for the hobbies we have; these things we own tell others something about ourselves. In fact, there is even a tradition that suggests that the ownership of property is necessary for the development of personhood. This tradition suggests that “Personality develops itself in its interaction with the world; without a sphere of property over which we exercise control, for example, moral responsibility is unlikely to develop” (Palmer 1990, 819).

Yet in practice, the notion of ownership does not translate quite so easily when we think about privacy in terms of access to information. Information, unlike a house or a car, can be held by multiple parties. In fact, this shared possession of information does not, by itself, mean that either party impacts (or even knows about) the other’s possession of it. A given set of information can still confer a lot of power and control upon those who possess it, and yet that possession can happen silently or invisibly. For these reasons, questions of intellectual property—that is, questions about who owns an intangible creation—are not only legally complicated but also have significant implications for questions of personhood.

Intellectual property law exists to protect a creator’s right to profit from the things that they have created in cases when that thing can be cheaply replicated, such as a poem, or a game, or a vaccine (Moore and Himma 2018). The existence of this area of law is, itself, evidence of the widely held belief that people own what they create and cultivate. But the ownership of something like this is not the same as what is considered to be personal information about us.

When we start to think about privacy in terms of property, some very interesting questions arise. For instance, do we consider our bodies to be our property? What about our talents, feelings, and experiences? If so, is information about our body also our property? Why or why not? There seems to be some basic consensus (e.g., Health Insurance Portability and Accountability Act [HIPAA] law) that our medical history is private information that an individual owns. But we also understand that this kind of information is necessary to share with medical professionals whom we rely on and trust.

If we own our bodies and our actions, then certainly we own these things in a different way from the objects we have in our home. But how is the information about our bodies something worth protecting, whereas information about our actions, behaviors and choices are open for anyone to collect and view? What about the things that we post about ourselves? What about the websites and other digital objects that we create?

These questions are becoming more and more important today when we have various forms of technologies that can be used to obtain, store, reproduce, and share information and data about ourselves and others. The traces that we leave are more abundant, and the transmission and sharing information about ourselves has a wider scope.

Warren and Brandeis defined privacy primarily as the “right to be let alone.” Their article, like many legal arguments today, suggested that the law needed to change to keep up with technological change. They noted that such constant updating was necessary, given the “expanding communication technology such as the development of widely distributed newspapers and multiply printed reproductions of photographs” (DeCew 2018). In order to understand privacy’s relationship to technology, it is important to keep in mind the legacy of Waren and Brandeis’s article and to think about other framings as well.

Over the years, various legal versions of privacy, typically tort or civil laws, have expanded what is covered and what is not covered by the interpretation of privacy as a protection over information. Prosser (1960, 389, qtd. in DeCew 2018) broadly outlines the four types of privacy that are covered under tort law in terms of informational privacy:

1. Intrusion upon a person and/or their private affairs.
2. Public disclosure of facts/information about an individual.
3. Placing someone in a false light publicly.
4. Appropriating one’s likeness for gain.

Rulings in recent years in the United States have also expanded the scope, concept, and reasoning behind informational privacy. Indeed, the courts have repeatedly stressed that the legal protection for privacy finds its source in the Fourth Amendment to the United States Constitution, which prohibits “unreasonable searches and seizures” (DeCew 2018). The reasoning following from the Fourth Amendment is again about control over information: you may have some information on your person that is necessarily protected from the government when the government has no reason to examine you.

Another operational definition of privacy comes from the training programs used for academic and professional researchers, for use in interacting with human subjects. These programs were established in the twentieth century as a direct response to research abuses that took place in the United States and Europe before and during World War II. These definitions were encoded into law with the National Research Act of 1974. The initial set of codes was adapted from the Nuremberg Codes at the close of World War II. From this training program we are given the following definition of both privacy and confidentiality (Hicks, n.d.).

• Privacy can be defined in terms of having control over the extent, timing, and circumstances of sharing oneself (physically, behaviorally, or intellectually) with others.
• Confidentiality pertains to the treatment of information that an individual has disclosed in a relationship of trust and with the expectation that it will not be divulged to others in ways that are inconsistent with the understanding of the original disclosure without permission.

These definitions for research practice have subsequently been used as the basis for definitions of privacy in a number of US regulatory requirements including HIPAA (Annas 2003).

In the second half of the twentieth century, a different legal interpretation of privacy emerged: decisional privacy, sometimes called the constitutional right to privacy (DeCew 2018). The notion of decisional privacy in US law extends from a reading of the Fourth Amendment against unreasonable search and seizure that differs from the one about information. In 1965, the Supreme Court heard the case of Griswold v. Connecticut (1965), which questioned whether a Connecticut law, restricting access to contraception, was legal under the Fourth Amendment. In the ruling, Justice William O. Douglas stated that the law violated the “right to marital privacy,” which laid the foundation for later rulings that protect a “zone of privacy” around marital relationships. The most controversial and most discussed implication of this decision was Roe v. Wade (1973) which, temporarily, made abortion legal in the United States (DeCew 2018). The fundamental idea is that there are certain decisions and spaces within which we are entitled to privacy and that the government should not be allowed to deny us this space. There is also an implicit idea here of the body as property: that we own ourselves and that what we do with ourselves is private.

This new reading of the Fourth Amendment in Griswold opened the door for many current discussions of privacy that include both control over personal information and control over our bodies and personal choices that are necessary for our concept of self (Kupfer 1987). The idea, generally, is that one can read the Fourth Amendment, and especially the ideas behind search and seizure, as protecting both information about oneself and the ability to make decisions, and that these rights should not be violated without some sort of due process. In a more modern reading, one can expand the definition of privacy to more abstract concepts of information—that is, all information about ourselves and what we do, wherever it may reside, is worthy of protection.

Perhaps the most comprehensive law passed with respect to informational privacy in recent years was the 2016 passage of the European Union’s General Data Protection Regulation (GDPR) (Voigt and Von dem Bussche 2017). This law defines a number of protections for EU citizens with respect to their data, independent of where that citizen is located or where the data is being processed. For example, the law includes the “right to be forgotten,” which means that any citizen may request that any company handling their data delete all data collected about them. The law also places strict requirements and penalties on companies that handle personal data of EU citizens and do not take proper care in safeguarding the data and/or do not share it in narrow and explicitly allowed ways.

The passage of GDPR has influenced and continues to influence privacy legislation around the globe. Understanding the complex legal and social history of the concept of privacy enables us to see that definitions about privacy are neither universal nor static. The United States, the European Union, and Asian countries have approached the legal protection of privacy differently. Within the United States there is a patchwork of laws, but generally there is only a limited system of formal protection (e.g., HIPAA), and personal data is readily available and easily collected for business or governmental use. This stands in stark contrast to the direction taken by the EU, especially with respect to the provisions in the GDPR which firmly place control of personal information in the hands of individuals, with strong protections accorded to individuals to control the flow and storage of information about themselves.

As we have seen, ideas of privacy have their roots in something fundamental that philosophers have struggled with for millennia. The modern definitions from which we sometimes (unknowingly) operate typically emanate from the developments of the legal tradition. In the United States especially, privacy has expanded beyond control over information and the right to be let alone, and now includes certain protections that protect the individual’s right to make choices about their lives and lifestyle (DeCew 2018). However, it’s important to realize that the law often lags broader social or cultural movements, the “policy vacuum” (Moor 1985), and this can also be observed in the legal history of privacy in the United States.

We have been revising and changing our definition of privacy for millennia, which reflects an ongoing commitment to the idea that there is something inherently valuable about privacy that is worth protecting. It has also changed with our conception of personhood. In the next section we discuss the value of privacy, and why privacy is necessary for one to develop a concept of self as a purposeful and self-determining agent.

So far we have discussed privacy primarily as a right that is typically associated with the ideas of property, bodily security, decision making, and especially personal information. This is arguably a legacy of the work of Warren and Brandeis, as most privacy discourse borrows their framing to describe privacy as a right—that is, a thing to which persons are entitled—rather than privacy as an interest—that is, a thing that people want and that benefits them in a morally significant way. Or, to use the language that we introduced earlier in the chapter, privacy might be construed as something that we have an interest in protecting because we value it (an axiological claim). We may value privacy because it is intrinsically valuable, or we may value privacy because it allows us to protect something we find especially valuable; for instance, personhood itself.

Matters of value, interest, and rights are deeply connected both conceptually and in everyday practice. But in this section, we consider more directly some reasons we might have an interest in protecting privacy for its own sake. In other words, we provide some answers to the question, why is privacy good? Why is it worth protecting?

Thus far, in our discussion of privacy, we have been emphasizing how valuable privacy is for a variety of reasons. But we should also consider how privacy is often in tension with other crucial goods and how it can create conditions for harm, abuse, and inequality.

Freedom from interference, whether from the government or the scrutinizing eyes of the public, seems like a good idea in principle, but it is an unfortunate fact that people sometimes use privacy to hide malicious behavior. This is especially the case in situations where people are underrepresented and/or are in vulnerable positions. For instance, from the perspective of feminist ethics, privacy is often questioned due to its historic use to hide physical, emotional, and financial harm (DeCew 2019).

Over the years, and especially in more modern settings, the value of privacy has been questioned, because privacy itself has been used as a shield to hide damaging actions (including domestic violence) and to preserve relationships of inequality.

Even if we grant that privacy is a right and something that is either good in itself or is a means to some other good end, this does not mean that it is something that will benefit all equally. In fact, it is quite easy to think of plenty of examples in which privacy is granted to some and not others. Both the parent-child relationship and the patient-client relationship contain not only an imbalance of power but also an imbalance of privacy.

Consider also situations in which someone we know is in need of care but refuses or is unable to reach out for help. For example, consider a scenario in which someone we know expresses suicidal thoughts, we may feel torn between sharing that knowledge with a licensed professional or someone else who can help and keeping that knowledge private. There are also times when people are called to make decisions for others for the sake of that person’s well-being, such as when a loved one begins showing signs of dementia.

These examples help us see that privacy is not a simple concept that can be applied to everyone equally, rather it is important to contextualize specific concerns about privacy. Returning to the questions we outlined in our discussion of utilitarianism in section 2.3.5 in chapter 2 might help to illustrate this point.

When we are considering the right to privacy, whose privacy rights are we protecting? Is anyone being left out, and if so, why? What is the value or the cluster of values that we are associating with privacy protection? Who has the most to lose if privacy is not protected? Who has the most to lose if privacy is protected? Should information about someone always be protected? Or is it OK to publish someone’s diary, for instance, after they die? If so, why is this OK?

Asking these questions and pushing ourselves to answer them helps us to sharpen our understanding about what privacy is and why it matters.

Before the beginning of the twentieth century, data about individuals was typically only collected through laborious, large-scale censuses (Wikipedia 2022a). As we have seen, beginning in the early 1900s and accelerating rapidly ever since, the ability to collect, store, transmit, and process data has rapidly increased. In addition to the ability to store it, we have also expanded what we consider as data. As discussed in chapter 3, what types of data are collected and how they are interpreted are neither natural nor obvious; rather, they result from viewpoints and framing, conscious or otherwise. Before the 1940s, the idea of tracking our shopping or TV watching habits was not considered data or even important (Wikipedia 2022b). As the twentieth century progressed, the use of behavioral data and simulation to understand populations, and the decisions they make, has continued to increase (Lepore 2020).

It used to require efforts on the level of governments to gather even basic information, such as the number of people living in a city. Now, tremendous volumes of data are available, and the collection and processing of this data is nearly real-time. It is now possible to track where a person was (e.g., through webcams, GPS data, and communications traces), what they did (e.g., through purchase records, check-ins, and Instagram posts), and how it affected them (e.g., through social media posts, biometric data from smartphones and activity trackers, and reviews). That data is often available to a number of different entities—human, corporate, enforcement-related, or otherwise—anywhere there is internet access. Much of this data is largely personal, extending beyond basic information to reveal a part of our personal identity, demographics, aesthetic preferences, and history of decisions made.

As we discuss in chapter 5, much of what has enabled this data collection is the rapid pace at which technology has intermediated many of our relationships: from the friends that we talk to on our smartphones to the movies that we watch on Netflix. This data has been called “the new oil,” and we are only starting to understand the implications of this data for business, the decisions we make, and what it means to us personally (Economist 2017). The continued increase in the availability of this data has fueled the modern revolution in AI and machine learning, which requires tremendous volumes of data to analyze, understand, and find patterns in our data (Kambhampati 2019).

Our own creation of data may be active, as when we post pictures online or fill out webforms, or it may be passive, happening without needing our awareness or express consent. Advertisers see which links produce incoming traffic and Netflix knows down to the second when you stopped watching that show. The data that we generate may be collected, accessed, used, or resold in ways that we did not fully understand (Madden 2014). According to Pew Research, most Americans feel that they have lost control of their personal data and what it is being used for. In this 2014 study, 91% of adults agreed that they had lost control of their data and more than 50% agreed that they did not understand how that data was going to be used by various companies (Madden 2014). In a 2018 study of users on Twitter, many were found to be unaware that their public posts could be collected and used by academic researchers, media companies, and many other entities. The majority of those surveyed felt that even researchers should get consent from users before collecting data, but that this feeling was highly contextual and depended on the type and scope of the research (Fiesler and Proferes 2018). However, in an updated Pew report, many were conflicted about the benefits of data collection and sharing (Auxier et al. 2019). For example, although many expressed concern about their map applications tracking their locations and saving it, many realized that the real-time traffic information, which is generated from the aggregated locations of everyone, presented a real value to them. Hence, in some instances users were happy to share personal data in exchange for the value received. However, in total, 81% of Americans thought that the potential risks of data collection by companies about them outweighs the benefits (Auxier et al. 2019).

How this data is collected and shared has been the subject of many studies, congressional oversight panels, and investigative journalism pieces. Dance et al. (2018) detailed a number of problems in the context of Facebook that allowed other companies to access more data than the company’s own privacy rules allowed. For instance, things such as private messages, photographs, and demographic information marked as for friends and family were used by Facebook to sell ads and were also provided to third-party companies including Netflix, Spotify, Yahoo, Amazon, and Bing for product targeting. Whereas this is ostensibly to allow for “instant personalization” across a number of sites, with the stated goal of providing more value to users, there have been a number of problems with these schemes. The Federal Trade Commission (FTC) investigated these data-sharing practices in 2011 and some were shuttered in 2014, but there are still many programs that share data that users have flagged as private (Dance et al. 2018).

The issue is bigger than any individual company that possesses our data. The collection processes and exchanges that they use allow them to build up fine-grained and detailed views of users, and their history of decisions, that is then used to fuel prediction models and advertisers (Economist 2017; Kambhampati 2019). Data is used not only by the companies whose products and services you use directly but also passed between numerous data brokers and aggregators, combined across websites, survey firms, political organizations, public records, and private companies; this data exchange creates profiles of individual users and their desires. A Vermont law requires all companies that buy and sell third-party data to register, providing a window into the process: customer management services such as Mailchimp, advertisers like Nielsen, and consultancies and businesses like ACXIOM, Oracle, and McKinsey, as well as financial data markets like Experian and other credit firms all buy, sell, and swap your data between themselves and other companies for uses including targeted marketing and parole risk prediction. These companies are third-party, meaning that they don’t have relationships with individuals and hence were required to register in Vermont, whereas Google, Facebook, and others would not need to register to pass on data (Christl 2017; Melendez and Pasternack 2019; Ram and Madhumita 2019). All this buying and selling, recirculating, and reconceptualizing of our data happens in ways that are not visible to us or might not make sense to us, even if we could know about it.

These collection practices raise concerns about the ownership of and access to our data, and how it translates to “public” and “private” space online isn’t at all clear, but, as Nissenbaum persuasively argues, this doesn’t mean that the online “world” is somehow a separate space, divorced from real life (Nissenbaum 2011). Online activity is deeply integrated into our social lives. “Answering questions about privacy online, like those about privacy in general, requires us to prescribe suitable, or appropriate, constraints on the flow of personal information” (Nissenbaum 2011, 14). Like offline activity, interactions online happen in a variety of social contexts. This is why it is a mistake to think about the “web” only in terms of a “commercial context where protecting privacy amounts to protecting consumer privacy and commercial information” (Nissenbaum 2011).

Reflecting on our discussion of personhood, we can start to see why losing control of our data feels so personal.