Cheating Statistics
What the General Social Survey, YouGov, the Kinsey Institute and the FTC actually show about infidelity in 2026 — traced from the first message to the moment it is discovered, with every figure linked to its source.
Who actually cheats more in 2026, men or women?
Among married Americans aged 25 to 54, women now report slightly more infidelity than men — 14% versus 11% — and the difference is not statistically significant. The gap everyone assumes did not narrow because women changed. It narrowed because men's reported rate roughly halved over three decades, and it halved across exactly the period in which an affair stopped requiring a place and started requiring only a screen. That is this page's argument, and the rest of it is the evidence.
The figure comes from Institute for Family Studies research director Wendy Wang, working with General Social Survey data — the longest continuous measure of American extramarital sex there is, run by NORC at the University of Chicago since 1972. In the 1990s, 21% of married men aged 25-54 said they had had sex with someone other than their spouse. By 2021-22, that was 11%. Women held broadly stable at 14% across the same span.
| Married adults aged 25-54 | 1990s | 2021-22 | Change (pp) | Relative change |
|---|---|---|---|---|
| Men reporting extramarital sex | 21% | 11% | -10 pp | -48% |
| Women reporting extramarital sex | 14% | 14% | 0 pp | 0% |
| Gender gap (men minus women) | +7 pp | -3 pp | -10 pp | reversed |
Source: Institute for Family Studies analysis of General Social Survey data (Wang), reported by Deseret News, August 2025. Relative change and gap columns are CatfishFinder calculations.
Our math: a drop from 21% to 11% is 10 percentage points in absolute terms but a 48% relative collapse — men's reported infidelity nearly halved. Most write-ups quote the two raw percentages and stop, which makes it read like a mild drift. It is not a drift. And because women stayed flat, the gender gap did not shrink to zero; it inverted, from men +7 points ahead to men 3 points behind.
Source: General Social Survey, analysed by the Institute for Family Studies
Broaden the lens past the 25-54 window and the older stereotype partly survives. Across all married adults, the GSS still finds roughly 20% of men and 13% of women admitting extramarital sex at some point in the marriage. Men's lifetime rate peaks around 26% in their seventies; women's peaks near 16% in their sixties. Those are cumulative lifetime figures for cohorts who married in a very different era — which is precisely why they disagree with the prime-age numbers above, and why quoting them as "the" cheating rate in 2026 is careless.
In any single year, by contrast, only about 4% of married adults report extramarital sex — a rate the Institute for Family Studies finds has stayed broadly flat since the 1990s. Hold on to that number. It becomes important later.
Where do affairs actually begin?
The two dominant starting points are the workplace and the private message thread — 39% of US workers say they have dated a coworker and 36% report a physical relationship with one, while nearly 60% of people who lost money to a romance scam in 2025 first made contact on social media. Neither of those is a dating app. The app is the venue people expect; it is not where most of the volume is.
The workplace numbers come from Resume Genius's 2025 Workplace Romance Report, a survey of 1,000 full-time US workers published in January 2025. Two details in it matter more than the headline. First, 72% of workplace romances are never disclosed to HR — Gen X are the most secretive at 86%, Gen Z the least at 52%. Second, 10% of US employees say they have matched with a coworker on a dating app, rising to 24% of Gen Z. Proximity and platform are not separate channels any more; they feed each other.
| Workplace romance behaviour | All US workers | Notable subgroup |
|---|---|---|
| Have dated a coworker | 39% | — |
| Physical relationship with a colleague | 36% | 45% of Gen Z |
| Romance never disclosed to HR | 72% | 86% of Gen X |
| Matched with a coworker on a dating app | 10% | 24% of Gen Z |
| Received uncomfortable romantic advances at work | 35% | 43% of women, 28% of men |
Source: Resume Genius Workplace Romance Report, survey of 1,000 US full-time workers, published January 2025.
Dating-app exposure is wide but shallower than the discourse suggests. A January 2026 SSRS survey of 2,012 US adults found 37% have ever used a dating site or app but only 6% currently do — 51% of 18-29s and 53% of 30-49s have ever used one, against 20% of adults aged 50 and over. Pew Research puts the ever-used figure at 30% in its 2023 benchmark. Both numbers are right; they are measuring slightly different frames, which is the theme of this page.
What the number hides: 37% have ever used a dating app but only 6% use one now — a ratio of about 6 to 1. That gap is the reason a partner's dormant profile is such a poor signal on its own. Six in seven people with an app history are not on one today, so an old account resurfacing in a search is weak evidence by itself. What carries weight is recent activity, a photo that appears elsewhere under a different name, or a second account under a handle you have never seen.
What counts as cheating, according to Americans?
Nine in ten Americans call sex with another person cheating, but 83% also count suggestive online messages and 73% count forming an intense emotional attachment — and only 23% consider every one of the ten behaviours tested to be infidelity. The definition is not a line. It is a gradient, and partners are frequently standing at different points on it.
YouGov surveyed 2,000 US adults in September 2022 on ten specific behaviours. The results:
| Behaviour | Share of Americans calling it cheating |
|---|---|
| Having sex with another person | 90% |
| Sending nude photos of yourself to another person | 90% |
| Sending suggestive online messages to another person | 83% |
| Lying about time spent with another person | 80% |
| Forming an intense emotional attachment to another person | 73% |
| All ten tested behaviours | 23% |
Source: YouGov, two surveys of 1,000 US adult citizens each, fielded 15-22 September 2022. Women rated every one of the ten behaviours as cheating more often than men did.
Source: YouGov survey of 2,000 US adults, September 2022
The same YouGov work produced the single most revealing pair of numbers on this page. Asked about their own conduct, 33% of Americans who have been in a monogamous relationship say they have cheated. Asked about their partners, 54% say they have been cheated on.
33% of Americans admit cheating in a monogamous relationship — but 54% say they have been cheated on. Reported victimisation runs 1.6x reported guilt.
Read this carefully: in a closed population, those two shares should sit much closer together than 33 and 54. Some of the 21-point gap is genuine — serial cheaters generate multiple victims, and one person can betray several partners across a lifetime. But no plausible amount of serial behaviour closes a gap that size. The rest is under-reporting, and it means the true rate almost certainly sits above every self-reported figure on this page, including the GSS ones. Every infidelity statistic you will ever read is a floor, not an estimate.
Infidelity also tracks demography more sharply than it tracks morality. From the same IFS analysis of GSS data on 25-54 year olds:
| Group | Reported extramarital sex | Comparison group |
|---|---|---|
| College graduates | 10% | 15% without a degree |
| Attend religious services regularly | 8% | 18% attending infrequently |
| Men in high-prestige jobs | 18% | 20% of unemployed men |
| Women in low-prestige jobs | 21% | 9% of women in high-prestige jobs |
| Divorced or separated adults (ever cheated) | 52% | 58% of women, 45% of men |
Source: Institute for Family Studies analysis of General Social Survey data (Wang), via Deseret News, 2025.
Note the direction reversal by sex on job status: prestige is mildly associated with more infidelity for men and markedly less for women. And notice the last row — among divorced and separated adults, women are the ones more likely to have cheated, 58% to 45%. That is the opposite of the folk model, and it is consistent with the prime-age reversal at the top of this page.
Is talking to an AI companion cheating, and how common is it?
One in seven partnered adults aged 18-30 now maintains a secret romantic attachment to an AI chatbot, and 61% of US singles say sexting or falling in love with an AI crosses the line — though only about a third call either behaviour cheating outright. This is the newest link in the mechanism, and it is the one the long-running surveys cannot see at all.
The prevalence figure comes from "Secret Soulmates", a 2026 study by Brigham Young University's Wheatley Institute and the Institute for Family Studies, based on 2,431 US adults aged 18-30 who were in real-life romantic relationships. The behavioural detail is the part worth reading twice.
| Among partnered US adults aged 18-30 | Share |
|---|---|
| Have a secret, frequent romantic attachment to an AI chatbot | 1 in 7 (14%) |
| Have experimented with an AI romantic companion at some point | 20-30% |
| Do not want their partner knowing the full extent of their use | 69% |
| Partner completely unaware of the AI involvement | 30% |
| Often role-play romantically or sexually with an AI | 13% |
| Often generate sexually explicit content with an AI | 11% |
| Use more than one AI romantic companion | 10% |
| Say it is easier to discuss feelings with an AI than a partner | 68% |
| Wish their real partner behaved more like the AI version | 50% |
Source: Secret Soulmates, Wheatley Institute (BYU) and Institute for Family Studies, 2026; n = 2,431 partnered US adults aged 18-30.
1 in 7 partnered 18-30 year olds have a secret romantic AI attachment — roughly 3.5x the annual rate of physical extramarital sex among married adults.
The catch: about 4% of married adults report extramarital sex in a given year, while 14% of partnered 18-30 year olds report an active secret AI attachment — roughly three and a half times as many. The two populations are not identical and the comparison is indicative rather than exact, so treat it as a magnitude, not a precise ratio. But the magnitude is the point. The behaviour that 73% of Americans would call cheating if it involved a human is now several times more common than the behaviour the national surveys have measured since 1972, and it appears in none of them.
Attitudes have not caught up either. Match and the Kinsey Institute's 14th annual Singles in America study, fielded in 2025 across 5,001 US singles, found 26% of singles now use AI for dating at all — a 333% year-on-year increase, with close to half of Gen Z already doing so. A different study — the Kinsey Institute and DatingAdvice.com's State of Us survey of 2,000 US singles, fielded April to May 2025 — produced the widely quoted 61% who say sexting or falling in love with an AI crosses the line. That 61% is a combined headline figure and it is worth unpacking, because the disaggregated numbers are much lower: asked about the two behaviours separately, 32% called sexting with an AI chatbot cheating and 29% said the same of an AI romance. The honest summary is that roughly a third of singles treat each behaviour as infidelity outright, a quarter already use AI in dating, and 69% of the young people using AI companions romantically are hiding the extent of it from their partner.
How do people actually find out about an affair?
Almost always through a device, and usually before any confession — 42% of daters say they have already used some form of technology to check on a partner they suspected. The phone has replaced the lipstick on the collar as the primary evidence of infidelity. This is the point in the mechanism where the digital shift cuts the other way: the same private channels that made affairs easier to start made them far harder to erase.
Every affair that runs through a messaging app, a second social account or an AI companion leaves artefacts: notification previews, login records, payment lines for subscriptions, timestamps that do not match the stated story, and photographs that exist in more than one place under more than one name. Pre-smartphone affairs left almost none of that.
Suspicion is also far more common than infidelity itself. In the Kinsey Institute and DatingAdvice.com State of Us survey of 2,000 US singles, 48% said they had suspected a partner of cheating at some point — 54% of women against 43% of men — and 42% of daters said they had used some form of technology, from AirTags to hidden cameras, to check.
Reality check: 48% have suspected a partner of cheating, while 33% admit to ever having cheated and 54% report having been cheated on. Suspicion sits almost exactly between confession and reported betrayal — which tells you suspicion is neither paranoid nor proof. It is roughly as accurate as a coin weighted slightly toward being right. That is the honest case for verifying rather than either dismissing the feeling or acting on it. A reverse phone check or a username consistency check answers a narrow, factual question: does this identity hold together across platforms. It does not answer whether someone is faithful, and no tool does.
A boundary worth stating plainly, because the industry around this topic often will not: installing spyware on another adult's phone, tracking their location without consent, or accessing their accounts is illegal in most US states regardless of what you suspect. Checking whether publicly posted information is consistent is a different act from surveilling a person. Stay on the right side of that line — a criminal charge is a worse outcome than an unanswered question.
What does infidelity cost a relationship?
Infidelity is the second most-cited contributing cause of divorce at 59.6%, but it is the single most common "final straw" at 24% — the thing that ends a marriage more often than it is the thing that broke it. That distinction is the most useful finding in the divorce literature and it is almost always flattened in summaries.
The source is Scott, Rhoades, Stanley, Allen and Markman (2013), published in Couple and Family Psychology, based on 52 divorced participants who had received premarital education. It is a small sample and should be read as such, but it is one of the few studies to separate contributing causes from terminal events.
| Reason for divorce | Cited as a major contributor | Cited as the final straw |
|---|---|---|
| Lack of commitment | 75.0% | — |
| Infidelity or extramarital affairs | 59.6% | 24.0% |
| Too much conflict or arguing | 57.7% | — |
| Getting married too young | 45.1% | — |
| Financial problems | 36.7% | — |
| Domestic violence | — | 21.2% |
| Substance abuse | — | 12.1% |
Source: Scott, Rhoades, Stanley, Allen and Markman, Reasons for Divorce and Recollections of Premarital Intervention, Couple and Family Psychology, 2013; n = 52. About 70% of participants identified a specific final straw.
One more finding from that study deserves its own line. Although 59.6% of individuals named infidelity as a contributing cause, only about 30% of former couples both named it. In most of these marriages, one person experienced the affair as the reason and the other did not — or did not know. Whatever else infidelity is, it is an event two people rarely remember the same way.
How does cheating overlap with catfishing and romance scams?
They share one mechanic — a person presenting an identity that does not match reality — and one of the two is measured precisely: US consumers reported $1.16 billion lost to romance scams in the first nine months of 2025 alone. The motive differs — concealment in one case, extraction in the other — but the verification problem facing the person on the other end is identical.
The scale of the deceptive-identity economy is documented far better than infidelity is, because money leaves records that feelings do not. In the first nine months of 2025, US consumers reported losing $1.16 billion to romance scams across 55,604 reports to the FTC's Consumer Sentinel Network — a 22% increase on the same period in 2024. In the third quarter alone, more than 11,200 people reported $398 million in losses, at a median of $2,218 each. Nearly 60% of people who lost money said the contact began on a social media platform. Across all scam types, the FTC reports more money lost to scams starting on Facebook than on any other platform, with WhatsApp and Instagram a distant second and third.
US consumers reported $1.16 billion lost to romance scams in the first nine months of 2025 — about $2,900 every minute, around the clock.
| Period | Reported romance-scam losses | Reports | Median loss |
|---|---|---|---|
| Full year 2023 | $1.14 billion | — | $2,000 |
| Full year 2024 | Sources range from ~$823m to ~$1.2bn | ~59,000 | $2,000 |
| Jan-Sep 2025 | $1.16 billion | 55,604 | — |
| Q3 2025 alone | $398 million | 11,200+ | $2,218 |
| 2025 annualised (our estimate) | $1.55 billion | ~74,000 | — |
Source: FTC Consumer Sentinel Network. The 2025 annualised row is a CatfishFinder projection from nine months of reported data and is not an FTC figure. See the methodology section on the 2024 discrepancy.
* CatfishFinder projection annualising the 1.16bn reported over the first nine months of 2025. Source: FTC Consumer Sentinel
Two calculations, and the second matters more. First, $1.16 billion across 273 days works out to about $4.25 million a day, or roughly $2,900 every minute, day and night. Second, and more useful if this is happening to you: divide $1.16bn by 55,604 reports and the average reported loss is about $20,900 — while the Q3 median is $2,218. The mean is more than nine times the median. That is a very long tail: most victims lose a few thousand dollars, and a small number lose life-altering sums. If someone is telling you their losses are unusual, the distribution says they are probably not.
These figures are themselves floors. The FTC counts reports, and shame suppresses reporting in this category more than almost any other. Our wider breakdown of that data sits in the catfishing statistics and online dating statistics pages.
Pew Research found that 52% of dating-app users have encountered someone they believed was a scammer — 63% of men under 50, and 44% of women of any age. Suspicion of fake identity is now the median dating-app experience, not an edge case. If you want to check whether a name, photo or handle holds together across platforms before you invest anything in it, our free identity consistency check is the place to start.
Why do cheating statistics disagree so wildly?
Because infidelity is measured entirely by asking people to confess, and because no two studies define cheating the same way — the credible figures on this page alone run from 4% to 54% — so the published range is not measurement error, it is definition drift. If you take one thing from this page, take this section. It is where the confusion lives.
1. "Ever cheated" and "cheated last year" are different questions. The GSS lifetime figure for married men is around 20%. The annual figure for married adults is about 4%. Both are GSS. Both are correct. A page that presents them side by side without saying which frame it is using is not reporting, it is decorating. Whenever you see an infidelity percentage, the first question is: over what period?
2. The measured population changes the answer. All married adults, married adults aged 25-54, everyone who has ever been in a monogamous relationship, and current dating-app users are four different denominators. That alone explains most of the spread between 11% and 33% on this page. It is why we have labelled the population next to every figure above rather than blending them into one headline number.
3. Self-report is a floor, and we can size the floor. Social-desirability bias means people under-report behaviour they are ashamed of. The 33%-admit versus 54%-were-cheated-on gap in the YouGov data is the cleanest available measure of that distortion. Anonymous online panels consistently return higher rates than interviewer-administered surveys for exactly this reason — which is one reason the GSS trend line, collected face to face for decades, should be read with care when it appears to show a decline.
4. The definition widened while the yardstick stayed narrow. This is the spine of this page. The GSS asks about sex with someone other than your spouse. It has effectively always asked that. Meanwhile 83% of Americans now count suggestive online messages as cheating, 73% count an intense emotional attachment, and 29% of singles count falling in love with an AI. That last figure is itself a lesson in framing: the same study is far more often quoted at 61%, because 61% is the share who said sexting or falling for an AI crosses the line when the two were offered together. Ask about them separately and you get 32% and 29%. The headline and the components are both accurate, and they support very different sentences — which is exactly how a single survey ends up cited as evidence for opposite claims. A survey instrument that measures one behaviour cannot detect a population migrating into four others. Men's reported extramarital sex fell 48% over precisely the decades in which private, deniable, non-physical channels became universal. We are not claiming the decline is entirely artefactual — attitudes and opportunity genuinely changed too. We are claiming that no honest reading of the GSS decline can rule the artefact out, and that almost every page reporting this number does exactly that.
5. Even hard financial data has definitional edges. The FTC's 2024 romance-scam total is quoted at roughly $823 million in some summaries and close to $1.2 billion in others. The most likely explanation is category boundaries: romance approaches that end in a fake investment platform can be logged as investment fraud rather than romance fraud, and different write-ups draw that line differently. When a source will not tell you which categories are inside the total, treat the total as a range.
What should you do if you think it is happening to you?
The single most useful finding for a worried person is that suspicion, at 48%, sits almost exactly between admitted cheating at 33% and reported betrayal at 54% — your instinct is meaningful evidence and it is not proof. That is an uncomfortable place to stand, and it is the honest one.
What the numbers on this page support:
- Do not rely on the old profile of who cheats. Among married adults aged 25-54, women report infidelity at 14% and men at 11%. Watching for one pattern means missing the other.
- The likeliest venue is closer than an app. 39% of workers have dated a coworker, 72% never told HR, and 69% of partnered under-30s using AI companions romantically hide the extent of it. Concealment is the constant; the platform is incidental.
- Agree on the definition before you argue about the behaviour. Only 23% of Americans call all ten tested behaviours cheating. A large share of infidelity conflicts are two people applying different rules in good faith, and that is a conversation, not an investigation.
- Verify identity, not fidelity. A phone consistency check or a username check can tell you whether an identity is coherent across platforms. Nothing online can tell you whether someone loves you. Confusing the two questions is how people end up spending money on tools that were never going to answer what they were actually asking.
- Money moving is the emergency signal. Reported losses ran near $2,900 a minute in 2025, and the average reported loss of about $20,900 is more than nine times the median. If a relationship that exists mainly through a screen has produced a request for funds, gift cards or crypto, the base rates stop being ambiguous.
The through-line: infidelity in 2026 has not obviously become more common. It has become less visible to the instruments that measure it. Men's reported extramarital sex halved over the same three decades in which every new channel for a secret relationship was invented, and the surveys that recorded that halving were never designed to count messages, emotional attachments, or a chatbot that 69% of its users are hiding from their partner. Read every cheating statistic — including all of ours — as the floor of a number nobody has fully measured.
Frequently asked questions
Do men or women cheat more in 2026?
Among married Americans aged 25 to 54, women report slightly more extramarital sex than men — 14% versus 11% — and the difference is not statistically significant, according to Institute for Family Studies analysis of General Social Survey data. Across all married adults including older cohorts, the lifetime figures still favour the traditional pattern at roughly 20% of men and 13% of women. The gap closed because men's reported rate fell by about 48% since the 1990s, not because women's rose.
What percentage of people cheat?
It depends entirely on the definition and the population. About 4% of married adults report extramarital sex in any given year (GSS), roughly 11-14% of married 25-54 year olds report it within the current marriage, and 33% of Americans who have been in a monogamous relationship say they have cheated at some point under a broader definition (YouGov, 2022). All three are self-reported and therefore floors.
Does talking to an AI chatbot count as cheating?
Many people say yes, but the number depends heavily on how the question is asked. In the Kinsey Institute and DatingAdvice.com State of Us survey of 2,000 US singles, 61% said sexting or falling in love with an AI crosses the line — but asked about the two behaviours separately, 32% called sexting with an AI chatbot cheating and 29% said the same of an AI romance. The behaviour is more common than many assume: a 2026 Wheatley Institute and Institute for Family Studies study of 2,431 partnered adults aged 18-30 found one in seven maintain a secret romantic attachment to an AI companion, and 69% do not want their partner knowing the full extent of their use.
How do most people discover an affair?
Through devices. Messaging apps, notification previews, subscription charges, login records and photographs reused across accounts leave records that pre-smartphone affairs never produced. Suspicion is common — 48% of singles say they have suspected a partner, 54% of women and 43% of men — and 42% of daters say they have used some form of technology to check, according to the Kinsey Institute and DatingAdvice.com State of Us survey.
Is infidelity the main cause of divorce?
It is the most common single trigger but not the most common underlying cause. In Scott et al. (2013), 75% of divorced participants named lack of commitment as a major contributing factor and 59.6% named infidelity — but infidelity was the most frequently cited final straw at 24%, ahead of domestic violence at 21.2%. Notably, only 31.3% of former couples both named infidelity as a cause.
Can I check whether someone is being honest about who they are?
You can check whether an identity is consistent — whether a phone number, email, username or photo appears elsewhere online under a different name or a conflicting profile. That is a factual question with a factual answer, and our free identity consistency check addresses it. Whether someone is being faithful is a different question that no online tool can answer. Accessing another adult's accounts or installing tracking software without consent is illegal in most US states.
How much money is lost to romance scams?
US consumers reported losing $1.16 billion to romance scams in the first nine months of 2025 across 55,604 reports to the FTC, up 22% year on year — about $2,900 every minute. Nearly 60% of people who lost money said contact began on social media. The Q3 2025 median loss was $2,218, while the average across all reports is closer to $20,900, indicating a small number of very large losses.
Sources
- Institute for Family Studies — Who Cheats More? The Demographics of Infidelity in America (General Social Survey analysis)
- Deseret News — 5 things to know about infidelity in the U.S. (reporting Wendy Wang / IFS GSS analysis, August 2025)
- YouGov — How many Americans have cheated on their partners in monogamous relationships? (n = 2,000, September 2022)
- Wheatley Institute (BYU) and IFS — Secret Soulmates: AI Romantic Companions and Young Adult Relationships (n = 2,431, 2026)
- Match and the Kinsey Institute — 14th Annual Singles in America Study (n = 5,001, June 2025)
- Kinsey Institute and DatingAdvice.com — State of Us: National Study on Modern Love and Dating in 2025 (n = 2,000 US singles, fielded April-May 2025; source of the 61%, 48% suspicion and 42% tech-checking figures)
- Resume Genius — Love on the Clock: 2025 Workplace Romance Report (n = 1,000 US full-time workers, January 2025)
- Scott, Rhoades, Stanley, Allen and Markman (2013) — Reasons for Divorce (Couple and Family Psychology, n = 52)
- Federal Trade Commission — data on losses to scams that start on social media
- FTC Consumer Sentinel Network Data Book 2024 (PDF)
- Pew Research Center — Key findings about online dating in the U.S.
- SSRS — The Public and Online Dating 2026 (n = 2,012, January 2026)
- Global Dating Insights — Romance Scams Outpace Other Fraud Types at $1.16 Billion (reporting FTC Consumer Sentinel data)
Derived figures on this page — relative change in reported infidelity, the confession-versus-victimisation gap, the per-minute and mean-versus-median loss calculations, and the 2025 annualisation — are CatfishFinder calculations from the sourced numbers above and are labelled as such wherever they appear.
Think Someone Might Be Catfishing You?
Run a free check to verify their identity in seconds.
Free Catfish Check