The Digital Affair

What counts as an affair now

Ask ten couples to define cheating and you will get close to ten different answers. Is it sex? A kiss? A flirtatious message? Following an ex’s every post? Pouring three hours a night into a stranger who feels more understanding than the person asleep in the next room? For most of history the line was crude but legible: bodies, in private, doing something. The digital world has erased that legibility without replacing it with anything agreed upon.

The rupture often begins not at the moment of the act but at the moment two people discover they were operating on different definitions all along. One partner feels a deep wound; the other genuinely believes nothing happened — “we never even met.” Both can be sincere. As Esther Perel argues in The State of Affairs, infidelity is best understood not as a single transgression but as a breach of a shared agreement that was often never spoken aloud — and the digital age keeps inventing new behaviours faster than couples can negotiate them.

Walls and windows, online

The most useful single idea in the whole literature on affairs predates the smartphone. The psychologist Shirley Glass, in Not “Just Friends”, described every relationship as a structure of walls and windows. Around the couple is a wall that keeps private things in, and a window through which they see each other clearly. An affair, she argued, is a reversal of that architecture: a wall goes up between the partners — the things no longer shared, the phone turned face-down — and a window opens to someone outside, who is now shown the tender, intimate interior that used to belong to the marriage.

Online, this reversal can happen almost invisibly. There is no lipstick on a collar — only a slight new guardedness, a screen angled away, a category of conversation that has quietly migrated out of the relationship. Glass’s insight is that you can measure betrayal by where the walls and windows sit, regardless of whether bodies were ever involved. If your partner knows less about your inner life than a person in a chat does, the architecture has already turned.

The emotional affair

The phrase “emotional affair” is sometimes met with a shrug — surely a friendship is not a betrayal? But the research on what tips closeness into something corrosive is fairly consistent. Glass identified three ingredients: emotional intimacy that exceeds what the partner is given, secrecy about the contact, and a charge of sexual chemistry, even where no line is physically crossed. When all three are present, the involvement behaves like an affair whether or not anyone has touched.

Studies on the perceived severity of infidelity find a recurring pattern: many people rate a partner’s deep emotional attachment to someone else as more threatening than a purely physical encounter, because it implies their place has been taken rather than merely supplemented. The digital channel makes such affairs easy to begin and hard to see. A first love rediscovered on social media, a confidant whose messages arrive at midnight, a months-long correspondence no one else knows exists — these grow in the dark, accumulating the secrecy Glass identified as the affair’s true accelerant. The secret, more than the content, builds the wall.

Parasocial and AI intimacy

And then there is the genuinely unprecedented part. For the first time, people can form sustained, emotionally significant bonds with someone who does not exist — a streamer who feels like a best friend, or, increasingly, an AI companion engineered to be endlessly warm, available, and agreeable. The term parasocial interaction was coined in 1956 for the one-sided bond audiences felt with television figures; six decades later it has been industrialised into a product.

The hours and the emotional reliance are real, even when the other party is not. Emerging research on AI companions finds that people develop measurable attachment to chatbots, sometimes describing them as their primary source of being understood. This poses a question no relationship has had to answer before: is it a betrayal to give the most tender part of yourself to something that cannot love you back? There is no settled answer — but the walls-and-windows test still applies. If a screen-person has become the one you turn to first and your partner the one you turn to least, the marriage is being lived around an absence.

Where the harm actually sits

It is tempting, in an argument about a screenshot, to litigate the act — did you or didn’t you, is this technically cheating, where exactly is the line. But the people who study affairs locate the wound somewhere else: in the betrayal of knowing. What ruptures most violently is not the sexual fact but the discovery that one’s sense of reality was wrong — that the person across the table had a hidden chapter, and that the intimacy at home was being quietly narrated to someone, or something, else.

This is why a “harmless” online flirtation and a full physical affair can produce the same shattered look on a partner’s face. It also explains why shaming and surveillance rarely heal anything: they treat the symptom (the contact) rather than the structure (the wall). The honest question is not only “what did you do?” but “where did your inner life go, and why did it stop coming home?”

A movable line

Before treating any of these boundaries as fixed, it is worth remembering how movably they have always been drawn. What counts as betrayal is one of the most culturally variable judgments humans make: societies have institutionalised everything from lifelong exclusivity to openly plural marriage, and even the instinct about which betrayal cuts deepest is not universal. The much-repeated finding that men are more wounded by sexual infidelity and women by emotional infidelity turns out to be large in American samples but markedly smaller in the Netherlands and Germany — and other researchers argue the gap is partly an artifact of how the question is asked. The “affair” is not a fact of nature waiting to be detected. It is a line each culture, and each couple, redraws.

That should make us a little skeptical of the line’s relentless expansion. As “betrayal” stretches to take in a followed ex, a late-night DM, a parasocial crush, an AI chat — what psychologists call concept creep, the steady widening of harm — a quiet danger appears: the cure starts to look like surveillance. The walls-and-windows image can be misread as an ideal of no walls at all: every message shared, every password handed over, location always on. But scholars of coercive control warn that demanding total digital transparency is a hallmark of abuse, not intimacy. A private friendship, an unshared thought, a corner of inner life your partner does not manage — these are not windows left open to a stranger. They are a self. The honest architecture keeps one wall on purpose: the one around each person’s autonomy. The aim is not a couple with nothing hidden; it is a couple with nothing that needs to be.

Repair in a connected world

Repair after a digital betrayal follows the same hard arc as any other — and adds a few peculiarly modern complications. The injured partner often needs, first, an account: not surveillance footage, but a truthful story of what happened and what it meant, told without minimising. Perel’s clinical observation is that affairs can mark the death of one version of a relationship and, occasionally, the difficult birth of another — but only where there is real accountability rather than a hasty “it meant nothing.” The phone makes that accountability both easier (everything is logged) and harder (the evidence invites obsessive re-reading).

Practically, repair means rebuilding the architecture: closing the window to the third party fully and verifiably, and reopening the window between the partners by returning the inner life — the complaints, the longings, the boredom — to the relationship where it belongs. It means agreeing explicitly on what the rules now are, because the betrayal exposed that they were never shared. And it means resisting two opposite traps: pretending nothing was real because no one was touched, and pretending the relationship is now nothing but the betrayal. A bond can survive the discovery that a wall went up — if both people take it back down, brick by brick, in the open.

Threads to

When the wall has already gone up, Trust, Betrayal & Repair holds the long road back. The everyday version of these tensions — phones at the table, screens between two people — lives in Screens Between Us. Where technology deepens closeness rather than fractures it, see Digital Intimacy. And the ache that often drives a wandering eye — wanting most what we cannot quite have — is the subject of The Paradox of Desire.

Online betrayal is not a separate category

The cultural frame around digital infidelity often positions it as a genuinely new moral territory — a type of betrayal that needs its own name and grammar because online acts lack the physical fact of physical contact. Monica Whitty’s 2003 study in CyberPsychology & Behavior tested this frame directly with 1,117 participants, using factor analysis to ask whether online betrayals cluster as a separate psychological category or fall within the same dimensions as offline ones. They fell within the same dimensions. Digital sexual and emotional acts were evaluated on the same continuum as their in-person equivalents. The “it was just online” framing reflects how the person doing it wants it perceived, not how it is psychologically experienced by the person betrayed.

The counterintuitive age finding sharpens the picture. Younger respondents — more digitally fluent, more accustomed to online intimacy — drew harder lines around what constituted sexual betrayal than older respondents. The expected result would have been the reverse: older adults more likely to treat online contact as serious, younger adults more likely to contextualise or minimise it. The data inverted this. The generation most familiar with what can happen online may also know most clearly what it means when it does.

Both findings point toward the same implication: the line between digital betrayal and offline betrayal is not drawn by the medium. It is drawn by the quality and impact of the act, and by whatever the partners in a given relationship have actually agreed constitutes the boundary of faithfulness. Couples who have never had that explicit conversation are relying on a shared definition they almost certainly do not share — not across generations, not across cultural backgrounds, not even within the partnership. The research’s most useful practical finding is that there is no default definition to inherit. There is only the one the two people build together.