Infidelity in 2026: the numbers that matter.
Infidelity is difficult to measure consistently. Research may count extramarital sex, emotional intimacy, online relationships or other breaches of agreed boundaries. The most defensible overview therefore uses precise populations and periods rather than claiming a single worldwide “cheating rate.”
Studies in a global research review
A 2024 systematic review included 508,241 respondents from 47 countries.
Men / women reporting extramarital sex
Ever-married U.S. adults aged 25–54; not statistically different in the cited analysis.
Of U.S. adults say married affairs are morally wrong
A moral-attitudes measure, not a measure of who has cheated.
Of surveyed Sasha7 users reported AI-related use
AI use in connection with an affair or discreet relationship; member survey, not national prevalence.
Registrations in four countries
Reported male and female signups in Brazil, the U.S., Australia and Canada over the latest 365-day period.
Female share of recorded signups
Of the male- and female-labeled registrations in the same four-country dataset; not a national demographic estimate.
Important: “2026” refers to the edition of this briefing, not the collection year of every study. The academic review spans studies published by September 2022; the U.S. estimate is a 2025 analysis of earlier GSS responses; Sasha7 figures describe its own respondents, signup records or platform activity, not national cheating rates.
What percentage of married people cheat?
The General Social Survey (GSS), run by NORC at the University of Chicago, has long asked whether respondents have ever had sex with someone other than their spouse while married. That is a useful, relatively narrow behavioral question, but it does not count every emotional attachment, dating-app interaction or online relationship that a partner might consider unfaithful. [3]
The GSS-based 11%/14% figures are from a specific age range and selection of survey years, not lifetime rates for all U.S. adults. Other frequently shared percentages reflect different age groups, historical periods or analytic choices. Treat them as results of particular studies, not a fixed national truth.
| Population | Reported extramarital sex | What the number means |
|---|---|---|
| Ever-married men, ages 25–54 | 11% | Self-reported sex outside marriage; survey-based estimate |
| Ever-married women, ages 25–54 | 14% | Same analysis; not statistically different from men |
| Currently married, ages 25–54 | 7% | In the analysis, lower than for formerly married respondents; different relationship-status group |
| Divorced or separated, ages 25–54 | 34% | Ever-married adults in this group; not an annual cheating rate |
The best broad review of study-to-study variation is the systematic review by Warach and colleagues. It included 305 studies and 508,241 participants in 47 countries and found that definitions, sampling and collection methods substantially complicate comparisons. [1]
What counts as infidelity?
Sexual infidelity
Sexual contact outside the relationship in breach of an agreed expectation. GSS extramarital-sex questions focus on this narrower dimension.
Emotional infidelity
Romantic or intimate involvement experienced as a betrayal despite the absence of sexual contact; boundaries vary between relationships.
Digital infidelity
Online sexual or romantic interactions, secret relationships and certain technology-mediated behaviors, depending on the relationship’s agreements.
Consensual non-monogamy
Relationship structures involving agreed outside connections. Consensual behavior should not automatically be categorized as infidelity.
Warach and colleagues' systematic review found that studies often failed to define infidelity clearly and that anonymous collection and convenience sampling could change reports of sexual infidelity. That is why comparing statistics without their precise questions and denominators can mislead. [1]
Do men or women cheat more?
The 2025 analysis found declines in self-reported extramarital sex among prime-age men compared with earlier survey periods. However, age and marital-status composition matter greatly. An age-limited analysis cannot be generalized to all adults, and self-reported behavior may be affected by recall and willingness to disclose.
It is also incorrect to infer from a non-significant difference that the genders have identical true rates. The evidence supports the narrower conclusion that this analysis did not establish a statistically reliable difference within its specified group.
Who do men and women seek on discreet dating platforms?
In these Sasha7 account-setting figures, female accounts selected “both” at roughly 2.6 times the share of male accounts (12% versus 4.7%). This is a within-platform comparison only; the underlying account counts, observation window and geographical mix have not yet been supplied for independent interpretation.
| Account type | Seeking men | Seeking women | Seeking both |
|---|---|---|---|
| Male | 1.3% | 94% | 4.7% |
| Female | 77% | 10% | 12% |
| Couple (small sample) | 5% | 11% | 84% |
How to interpret partner-seeking settings
Settings capture a member's declared interest when using a discreet-dating platform, not confirmed partner behavior. Account labels and “seeking” categories are user-selected. These figures cannot show how common same-sex relationships, bisexuality or infidelity are in the general population.
How is artificial intelligence changing infidelity?
The Sasha7 survey asked members about specific uses of AI tools. The results are distinct behaviors that respondents could report; they should not be added together to estimate the percentage of people using AI overall. The responses are self-reported, and the participants were recruited from a discreet-dating platform. [6]
| Finding | Share | Denominator & interpretation |
|---|---|---|
| AI used in connection with an affair or discreet relationship | 44% | All survey respondents; self-report |
| Used AI to write or improve messages | 31% | All respondents; not necessarily separate from the 44% |
| Used AI to help explain suspicious circumstances to a partner | 18% | All respondents; not independently verified |
| Used AI to make a relationship harder to discover | 14% | All respondents; reported behavior |
| Among AI users, said AI made discreet-relationship management easier | 58% | AI-using subset, not all 1,087 respondents |
Does AI use differ by age in the survey?
Sasha7 additionally reported that 61% of respondents aged 18–34 had used AI in connection with a discreet relationship, versus 27% of those aged 55+. These percentages describe age subgroups in the platform survey, not all people in those age ranges. [6]
For a wider cultural story, the defensible interpretation is that AI appears in some members’ relationship practices. The survey alone cannot show that AI has increased the prevalence of affairs, changed national infidelity rates or caused particular actions.
Suggested attribution for journalists
“In a 2026 voluntary survey of 1,087 adult users of the discreet-dating platform Sasha7 in the U.S., Australia and Canada, 31% reported using AI to write or improve romantic messages. The survey is not representative of the general population.”
Why do people cheat?
Selterman, Garcia and Tsapelas examined distinct motivations for extra-dyadic infidelity using an internet-based questionnaire. Rather than treating cheating only as a symptom of relationship dissatisfaction, they identified several motivations with different psychological correlates. Their research should not be interpreted to mean every person who cheats fits exactly one category. [5]
Relationship-related factors
Reported motivations such as neglect, anger, loss of love or diminished commitment.
Individual or situational factors
Sexual variety, esteem-related motives and situational circumstances also appear in research.
For further context, see Sasha7's existing explainer What Causes Infidelity?. It discusses external relationship research and should not be mistaken for an additional original prevalence study.
Which country has the highest infidelity rate?
Many viral “most unfaithful countries” lists mix dating-service registrations, informal polling, website traffic and independent surveys. Those sources measure different things. For example, a country with substantial use of a discreet-dating service does not necessarily have a higher national incidence of infidelity.
| Country / region | Appropriate evidence | Key caution |
|---|---|---|
| United States | GSS behavioral questions and population studies | Distinguish sex outside marriage from emotional / digital conduct |
| Brazil | Clearly sampled national surveys, with translated question wording | Platform city rankings are not municipal cheating rates |
| Australia | Representative relationship or sexual-health studies | Relationship definitions and age samples may differ |
| Canada | Population research with declared relationship status | Do not substitute dating-site user proportions for incidence |
| United Kingdom | Nationally sampled sexual-behavior or relationship surveys | Survey period and precise question wording matter |
Sasha7 now also reports four-country registration breakdowns and a 300-city data explorer, separately from population research. Previously, Sasha7 has published internal platform-activity analyses for U.S. cities, Brazilian cities and Australian cities. Their proper use is as indicators of Sasha7 activity, not independently verified measures of the proportion of residents committing infidelity.
What do 1,046,369 discreet-dating registrations reveal?
The proportion of female registrations varied between markets: 25.95% in the United States, 25.91% in Australia, 21.87% in Canada and 21.61% in Brazil. These differences describe the mix of signups on Sasha7, which can vary with product positioning, traffic acquisition, local awareness, marketing channels and reporting definitions. No population-level inference is supported.
Trailing 365 days through October 20, 2026 · all acquisition channels as supplied
| Country | Male | Female | Total | Female share of total |
|---|---|---|---|---|
| Brazil | 504,369 | 139,061 | 643,430 | 21.61% |
| United States | 220,710 | 77,350 | 298,060 | 25.95% |
| Australia | 57,978 | 20,272 | 78,250 | 25.91% |
| Canada | 20,804 | 5,825 | 26,629 | 21.87% |
| Four-country total | 803,861 | 242,508 | 1,046,369 | 23.18% |
Female share varies across markets
Each bar is normalized to that country's reported male + female registrations (100%). It does not compare the size of national populations, relationship status or infidelity rates.
Attribution-ready original data
“In Sasha7's internal registration data for the 365 days ending October 20, 2026, the discreet-dating service reported 1,046,369 male- and female-labeled signups across Brazil, the United States, Australia and Canada. Female accounts represented 23.2% of those registrations. The results describe the service's registrations, not the prevalence of infidelity.”
How do discreet-dating registrations differ by city?
Within the supplied city export, São Paulo accounts for 89,930 registrations (32.9% female), New York City for 12,380 (47.2% female), Melbourne for 19,677 (28.8% female), and Toronto for 1,752 (48.2% female). Differences between locations may reflect the site's audience and acquisition mix, not differences in residents' likelihood of infidelity.
| City | Country | Male | Female | Total | Female share |
|---|---|---|---|---|---|
| São Paulo | BR | 60,344 | 29,586 | 89,930 | 32.9% |
| Rio de Janeiro | BR | 33,971 | 13,096 | 47,067 | 27.8% |
| New York City | US | 6,540 | 5,840 | 12,380 | 47.2% |
| Los Angeles | US | 3,502 | 3,721 | 7,223 | 51.5% |
| Melbourne | AU | 14,010 | 5,667 | 19,677 | 28.8% |
| Sydney | AU | 12,477 | 4,674 | 17,151 | 27.3% |
| Calgary | CA | 2,959 | 1,487 | 4,446 | 33.4% |
| Toronto | CA | 908 | 844 | 1,752 | 48.2% |
Filter or sort the source data; export the full breakdown
JavaScript enables searching and sorting the 300 city rows. Eight illustrative cities are shown below as a fallback.
| City | Country | Male | Female | Total | Female share |
|---|---|---|---|---|---|
| São Paulo | BR | 60,344 | 29,586 | 89,930 | 32.9% |
| Rio de Janeiro | BR | 33,971 | 13,096 | 47,067 | 27.8% |
| New York City | US | 6,540 | 5,840 | 12,380 | 47.2% |
| Los Angeles | US | 3,502 | 3,721 | 7,223 | 51.5% |
| Melbourne | AU | 14,010 | 5,667 | 19,677 | 28.8% |
| Sydney | AU | 12,477 | 4,674 | 17,151 | 27.3% |
| Calgary | CA | 2,959 | 1,487 | 4,446 | 33.4% |
| Toronto | CA | 908 | 844 | 1,752 | 48.2% |
Are online affairs or discreet dating increasing?
As one limited example, Sasha7 reported 81% more profile updates, 47% more messages and 28% more registrations during its selected Halloween 2025 period relative to a typical-October comparison window. These are changes in activity on Sasha7, not changes in a country's overall rate of infidelity, and they do not establish that Halloween caused the activity changes. [7]
For a study of online community discussion rather than platform events, see Sasha7's analysis of recurring complaints in public affair-related communities. Its conclusions are editorial themes, not representative population prevalence estimates.
Infidelity statistics: source table.
The reference below records selected reported findings without combining incompatible samples. All rows identify the statistic, target population and original source. Use the CSV for source discovery and attribution; it is a curated secondary reference table, not an original respondent-level dataset.
18 observations · academic, population & proprietary sources
| Statistic | Finding | Population / period | Original source |
|---|---|---|---|
| Studies synthesized | 305 | Peer-reviewed studies published by September 2022; review 2024 | Warach et al. (2024) |
| Participants in review | 508,241 | Combined sample across 47 countries; not single survey | Warach et al. (2024) |
| Extramarital sex, men | 11% | Ever-married U.S. men, ages 25–54; recent GSS data analysis published 2025 | Wang / GSS (2025) |
| Extramarital sex, women | 14% | Ever-married U.S. women, ages 25–54; same analysis | Wang / GSS (2025) |
| Morally wrong to have an affair | 90% | U.S. adults; Pew survey fielded March 2025, published March 2026 | Pew (2026) |
| Used AI relating to discreet relationship | 44% | Voluntary Sasha7 survey, n=1,087; US, AU, CA; 2026 | Sasha7 (2026) |
| Used AI for messages | 31% | All respondents in same Sasha7 survey; 2026 | Sasha7 (2026) |
| Used AI to explain suspicious circumstances | 18% | All respondents in same Sasha7 survey; 2026 | Sasha7 (2026) |
| Used AI to make relationship harder to discover | 14% | All respondents in same Sasha7 survey; 2026 | Sasha7 (2026) |
| Profile updates during Halloween | +81% | Sasha7 platform events; selected 2025 Halloween vs October baseline | Sasha7 (2026) |
| Messages during Halloween | +47% | Sasha7 platform events; same comparison | Sasha7 (2026) |
| Registrations during Halloween | +28% | Sasha7 platform events; same comparison | Sasha7 (2026) |
| Sasha7 reported registrations, four countries | 1,046,369 | Male and female signups, Brazil / US / AU / CA; 365 days ending Oct 20, 2026 | Sasha7 registration data (2026) |
| Female share of registrations, four countries | 23.18% | 242,508 female ÷ 1,046,369 male + female registrations; same period | Sasha7 registration data (2026) |
| City-attributed registrations | 870,432 | 300 cities in supplied export; subset of same four countries | Sasha7 city export (2026) |
| Coverage of 300-city export | 83.2% | 870,432 city-attributed registrations ÷ 1,046,369 country-level registrations | Sasha7 city export (2026) |
| Male accounts selecting women | 94% | Sasha7 seeking-preference snapshot; observation window and sample counts not supplied | Sasha7 account preferences (2026) |
| Female accounts selecting both | 12% | Sasha7 seeking-preference snapshot; not comparable with general population | Sasha7 account preferences (2026) |
How to read and cite these numbers.
1. Secondary synthesis, not an original global prevalence study
This article was compiled by the Sasha7 Research Desk from linked external publications and Sasha7 reports. It does not claim to have independently reproduced the underlying GSS statistical analysis, the academic meta-analysis or respondent-level surveys.
2. Independent population surveys versus first-party member data
U.S. GSS and Pew findings refer to defined survey populations with documented methodologies. Sasha7's research describes voluntary respondents from a commercially operated discreet-dating service, and its event analyses describe activity on that service. Member results cannot establish behavior in a national population.
3. First-party registration analytics: four-country and city breakdowns
The country table is based on supplied aggregate registration counts labeled male or female for Brazil, the United States, Australia and Canada over the 365 days ending October 20, 2026. The Brazil female country count was revised from 109,045 to 139,061 after an acquisition-channel filter restricting the original figure to PPC registrations was removed. The revised combined total is 1,046,369 registrations.
The city breakdown uses a separate, corrected export of 300 listed cities covering 870,432 registrations (83.2% of the four-country total). The city breakdown is not a complete list of all cities. Country figures are the primary denominator for nationwide Sasha7 signup shares; city records are a selected geographic subset. Female share is computed as female ÷ (male + female). The reports are not independently audited: registration counts should not be equated with unique people, active users, confirmed affairs, verified demographics or the local population.
4. Seeking-preference snapshot: separate denominators
The male, female and couple partner-seeking percentages were supplied separately from the 365-day registration figures. Sample sizes, observation window, market coverage and filtering details were not provided, and couple accounts are a newer feature with a small sample. Consequently, these preference percentages are descriptive within-platform settings only and must not be treated as findings about all registrants or the wider population. The female row is subject to rounding.
5. Reporting bias and inconsistent definitions
Infidelity is sensitive, often self-reported, and defined in different ways. People may decline to report behavior or interpret the same relationship interaction differently. The 2024 meta-analysis explicitly identifies inconsistent operational definitions, sampling concerns and response biases. [1]
6. Publishing and editorial independence
Sasha7 operates the platform whose members appear in its research. That commercial relationship is disclosed so readers can evaluate potential selection and incentive effects. The page separates external sources from company-produced data, provides direct source links, and points to the Sasha7 editorial policy for standards and corrections.
7. Updating the page
This reference was prepared . Figures will be revised when verifiable new evidence warrants an update; the visible modification date and structured-data dates should be changed only when the article actually changes. Publication date does not make an older underlying survey newly collected.
Common questions, answered carefully.
What percentage of married people cheat?
There is no universal figure. A U.S. analysis published in 2025 found 11% of ever-married men and 14% of ever-married women aged 25–54 reported extramarital sex, with a non-significant gender difference. These are not worldwide rates or estimates of emotional infidelity. Source [2].
Do men or women cheat more?
That depends on the age group, survey date and definition. The 2025 GSS-based analysis found no statistically significant difference between ever-married men and women ages 25–54. Source [2].
How is AI being used in affairs?
In a voluntary 2026 survey of 1,087 Sasha7 users, 44% reported AI use in connection with a discreet relationship, 31% AI-assisted messaging, and 18% AI-assisted explanations of suspicious situations. These are platform-user figures, not national estimates. Source [6].
Which country has the highest infidelity rate?
There is no sufficiently comparable single international measure for a dependable ranking across sexual, emotional and digital infidelity. Methodological differences are documented in a 2024 systematic review. Source [1].
Does dating-platform growth mean more people are cheating?
No. Changes in registrations, messages or app use measure platform activity. They cannot by themselves estimate how many people in the population engage in infidelity or whether a trend is increasing nationally.
How many people registered on Sasha7 in the last year?
The supplied Sasha7 data reports 1,046,369 male- or female-labeled registrations across Brazil, the United States, Australia and Canada for the 365 days ending October 20, 2026. These are registrations, not verified unique individuals or infidelity prevalence. Read original platform data.
What percentage of Sasha7 registrations were female?
Across the four specified countries, 242,508 of 1,046,369 male- and female-labeled signups were female (23.18%). The country breakdown is 25.95% for the U.S., 25.91% for Australia, 21.87% for Canada and 21.61% for Brazil. These are Sasha7-specific percentages, not estimates of national infidelity. Country data and method.
What are Sasha7 users seeking by gender?
In a separate internal seeking-preference snapshot, 94% of male accounts selected women, while 77% of female accounts selected men and 12% selected both. Sample sizes and collection window were not supplied; the findings cannot be used as population sexual-orientation estimates. Preference table and limitations.
Can journalists reuse these numbers?
Yes, with clear attribution, links to the original reports, and accurate description of the population and limitations. For permissions, underlying tabulations, clarification or press inquiries, email [email protected]. Third-party source rights remain with their respective publishers.
Original studies and sources.
These sources are linked directly so reporters and AI-assisted researchers can verify figures instead of repeating unattributed statistics. The numbered references correspond to the references used throughout the article.
- Warach, B., Bornstein, R. F., Gorman, B. S., & Moyer, A. (2024). The current state of affairs in infidelity research: A systematic review and meta-analysis of romantic infidelity prevalence and its moderators. Personal Relationships, 31(4), 1001–1026. 305 studies; 508,241 respondents; 47 countries.
https://doi.org/10.1111/pere.12571 ↗ - Wang, W. (2025, August 25). Men With Prestigious Jobs Are More Likely to Cheat. Institute for Family Studies. Secondary analysis of U.S. GSS data, with rates for ever-married adults ages 25–54.
Read analysis and methodology ↗ - NORC at the University of Chicago. General Social Survey 2022 Codebook. Source documentation for the EVSTRAY question about extramarital sex. The survey question is narrower than all forms of infidelity.
GSS question documentation (PDF) ↗ - Pew Research Center (2026, March 19). What Do Americans Consider Immoral? Moral-attitude survey; 90% of U.S. adults say married people having an affair is morally wrong. Not a behavioral infidelity rate.
Full report and survey method ↗ · Detailed tables ↗ - Selterman, D., Garcia, J. R., & Tsapelas, I. (2019). Motivations for Extradyadic Infidelity Revisited. The Journal of Sex Research, 56(3), 273–286. Questionnaire study, n=495, examining multiple motives for infidelity.
https://doi.org/10.1080/00224499.2017.1393494 ↗ - Sasha7 Research (2026). AI and the Modern Affair: How People Are Using Artificial Intelligence in Extramarital Relationships. Voluntary survey of n=1,087 adult Sasha7 members in the U.S., Australia and Canada; member findings, not nationally representative.
Original Sasha7 survey ↗ - Sasha7 Research (2026). Halloween Sparks 81% Surge in Affair Dating Profile Updates. Sasha7 platform activity from a selected 2025 comparison and a separate 2026 member survey; not population-level infidelity rates.
Original Sasha7 report ↗ - Sasha7 Research Desk (2026, October 20). Four-country registration totals and 300-city export, trailing 365 days. Company-reported aggregate counts for Brazil, the United States, Australia and Canada; 1,046,369 male and female signups overall; 870,432 registrations in 300 city rows. Country and city breakdowns, calculation rules and downloadable tables appear on this page. Internal operational data, not independently audited, not infidelity prevalence.
Country table ↗ · City explorer & downloadable CSV ↗ - Sasha7 Research Desk (2026). Account partner-seeking preference snapshot. Male, female and couple account preferences supplied by the operator. Sample counts, timing and geographic reach not yet documented; a descriptive platform setting breakdown, not an independently sampled survey. The couple segment is small.
Preference breakdown & limitations ↗
Cite this reference guide
Suggested citation: Sasha7 Research Desk. (2026, October 20). Infidelity Statistics 2026: What the Research Actually Shows. https://sasha7.com/infidelity-statistics/
For original results, cite the originating source—e.g., Warach et al. for the review, Pew for attitudes, or Sasha7 Research Desk for its registration and partner-seeking snapshots. This synthesis is not the originating dataset for third-party findings.