Every major dating app now advertises safety tech — photo checks, AI scam-detection, background checks. Below is what each one actually publishes, and the one thing none of them can do: tell you whether a real, verified person is being honest with you.
Dating-app verification confirms that a profile is a real, live person who matches their photos — it defeats stolen-photo catfishing. It does not confirm that the person is honest, and it stops working the moment the chat moves off the app. So the two rules no feature can enforce are the ones that matter: keep the conversation on-platform until you've met, and never send money or "invest" for someone you met online.
Match Group app
Bumble Inc.
Match Group app
Match Group app
Match Group app
Match Group app
Grindr Inc.
WINK TECH LIMITED — the college-marketed newcomer
Feeld Ltd — the open-minded / non-monogamy segment
Pure App — the anonymous hookup segment
Perry Street Software — the gay / queer hookup segment
Read the entries above and a pattern appears. The apps have quietly built genuinely useful tech — a live selfie compared to your photos, AI that flags a scam-shaped message, a badge that says a human is behind the profile. All of it attacks the same old problem: the catfish using someone else's pictures. And all of it answers the same question — is this person real? — while leaving the more dangerous question untouched: is this person honest? That gap has a price tag: the FBI's Internet Crime Complaint Center recorded $929 million in confidence-and-romance-scam losses across 23,159 complaints in 2025.
One structural note: Tinder, Match.com, Hinge, OkCupid and Plenty of Fish are all Match Group apps and share much of the same verification infrastructure — so their photo/selfie checks work alike, and the discontinued Garbo background-check partnership affected all of them.
That gap is not an oversight; it's unsolvable by software. The costliest dating-app scams — romance scams and pig-butchering investment cons — are frequently run by real people, on real video, using their own faces, following a script. A selfie check waves them straight through. What gives them away is never the profile; it's the behaviour — the excuses that avoid a real meeting, and the moment the conversation bends toward money or a trading platform.
So use the features — verify your profile, keep the chat where the moderation lives, report anything off. But remember what they can and can't do. If something feels off about someone you've met on an app, run it through our romance-scam checker, which reads the behaviour the badge can't. Reviewed August 2026.
The index asks one question of each app, and answers it only from what the platform itself publishes: what does this app actually verify, and what does it explicitly not do? Every row is sourced to the platform's own safety, help or newsroom page, or to a named authority. Where a claim could not be confirmed on a platform-owned page, it is omitted rather than guessed.
That discipline matters here, because this category is full of features that sound like safety and are not. Verification confirms that a live human face matches the profile photos. It does not verify identity, intent, or criminal history — and several apps say so themselves. So the index records what each tool actually proves, not what its badge implies.
The single most important column is the one that is empty for every app: background checks. Tinder offered them through Garbo from 2022; that partnership ended in 2023, and across all 12 apps no replacement is live. We also note the structural point once: Tinder, Match.com, Hinge, OkCupid and Plenty of Fish are Match Group apps sharing much of the same verification infrastructure, so they are not 12 independent safety regimes.
New in v2.0.0, and the sharpest finding in this release: of 12 apps, exactly 1 documents true two-factor authentication on the account itself. That is eharmony, which emails a 4-digit code when a login is attempted from an unrecognised device. Everywhere else the SMS or email code is the login, which is not the same thing: a code that replaces a password protects nobody who has already been SIM-swapped or had their inbox taken. Grindr is the only app publishing a PIN plus biometric lock on the app itself, which answers a different threat, somebody holding your unlocked phone.
Only 4 of 12 check a government ID at all, and where they do it is usually narrower than it sounds: Hinge's is an age check that becomes compulsory once prompted, and PURE's exists to prove age in the United Kingdom rather than to prove who anyone is. A photo badge and an ID check are different claims, and most apps only make the first.
What happens to the face afterwards varies more than anything else here, and the real split is not between companies but between kinds of check. A one-off verification is cheap to forget: OkCupid purges its facial-geometry template within 24 hours, Hinge deletes Selfie Verification data inside the same window, and Bumble's ID partner Veriff deletes within 72 hours. The newer biometric age checks keep far more, and keep it far longer. On Tinder and on Hinge alike, the FaceMap and the FaceVector derived from your video selfie are retained for the lifetime of the account and deleted within 30 days of closure, with audit images held for 90 days past closure on Tinder and up to a year if the account was banned. Bumble keeps biometric scans up to three years after your last interaction, and Grindr retains data for as long as the account is active. And 3 of 12 still publish no retention period at all, which is the answer you get for handing over your face and then asking what happens to it.
Last reviewed August 2026. Platforms ship new tooling often — each row links its source so any claim can be checked, and corrected, at source.
This index is free to cite, with credit (CC BY 4.0). The one-line finding: not one of the 12 major dating apps publishes a background-check programme, and only 1 of 12 offers two-factor authentication on the account itself.
The dataset is archived with a permanent DOI, so it stays citable and resolvable independently of this site: 10.5281/zenodo.21440084 (CSV + JSON + methodology, 2.2.0). That is a concept DOI, so it always resolves to the newest version rather than pinning to one deposit. To cite this exact revision instead, use 10.5281/zenodo.22227448.
Suggested citation: Tutela Digitalis (2026). The Tutela Digitalis Dating App Safety Index (Version 2.2.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.21440084
Need the underlying data, a correction, or a quote? Email press@tuteladigitalis.com. We publish corrections.
The platforms themselves are legitimate, publicly-run companies, and most now publish real safety tech — photo or selfie verification, AI moderation, in-app reporting. What they cannot make safe is the other person. Every verification feature answers one question — is this a real, live human? — and none of them answer the question that actually matters: are this person's intentions honest? A verified profile can still belong to a scammer, because verification confirms a face, not a motive. So the safe way to use a dating app is to treat its safety features as a floor, not a guarantee, and keep the two habits no app can enforce for you: stay on-platform, and never send money.
Photo or selfie verification (Tinder, Bumble, Hinge) compares a live selfie to the profile photos to confirm the person is real and matches their pictures — it defeats stolen-photo catfishing, but nothing more. A background check (offered by some Match Group apps through a third party) searches public records for a criminal history under a name you supply. Neither tells you whether the person is honest with you, and neither can see what happens once a conversation moves off the app. A romance scammer can pass a selfie check with their own face and still be running a script.
No — verified means real, not trustworthy. Verification badges are worth having (they filter out the laziest stolen-photo fakes), but the most damaging dating-app scams — romance scams and pig-butchering investment cons — are often run by real people using their own faces, sometimes themselves coerced. The reliable signals aren't the badge; they're behaviour: will they hold a sustained live video call, will they meet in person, and do they ever steer the conversation toward money, crypto, or an 'investment opportunity'? A refusal to video-call plus any money ask outweighs any badge.
Because the app's safety tech only works inside the app. Reporting, AI message-scanning, and moderation all stop at the moment you switch to WhatsApp, Telegram, Google Chat or text — which is exactly why an early, insistent push to move off-platform is one of the strongest scam signals there is. A genuine match has no reason to rush you off the app in the first hours. The defence is simple: keep talking where the safety tools live until you have met in person, and treat 'let's move to WhatsApp' from a stranger as a yellow flag.
Any request for money, or any pitch to 'invest' — no matter how the story is dressed. Whether it's a stuck-abroad emergency, a customs fee, a medical bill, or a can't-miss crypto or trading platform they want to 'teach' you, a request to send money or move funds to an investment is the line a genuine romantic interest never crosses. Pig-butchering scams add a cruel twist: the trading dashboard shows fake profits to pull you deeper, and the 'withdrawal fee' at the end is the final squeeze. If money enters the conversation, stop — and run it through our romance-scam checker.
No — it's a safety reference, not a ranking. We're not scoring the apps against each other or recommending one; we're documenting what each platform publishes about its own anti-scam and verification tools, so you can see what protection you actually have and where the gap is. Every entry is sourced to the platform's own safety or help page and reviewed periodically. Features change, so confirm current details with the app itself.
Trust them for exactly what they claim — and nothing more. Hinge's Selfie Verification pairs a Liveness Check with 3D Face Authentication and deletes the facial-geometry data within 24 hours; Hinge itself says plainly that verification “doesn't guarantee the safety of a particular user.” Bumble's Deception Detector combines AI with human moderation to catch fake, spam and scam profiles before members see them — Bumble reported it blocked 95% of such accounts automatically. Both are real, useful systems, and both answer the same narrow question: is this a live human whose face matches the photos? Neither evaluates honesty. A verified Hinge or Bumble profile can still be a real person running a script — so the trust test stays behavioural: a sustained live video call, a real-world meeting, and zero tolerance for any money or investment talk.
No dating app is structurally “secure” against a scammer using their own real face — but the published verification floors do differ. The strongest identity checks in this index are the biometric ones: Tinder (Photo Verification plus the Face Check rollout, required for new users in several markets), Hinge (Selfie Verification with a Liveness Check and 3D Face Authentication), Bumble (AI photo verification plus Deception Detector moderation) and, among the newcomers, Duet's enforced selfie face check. At the other end, eharmony and Grindr rely on SMS verification, which only proves control of a phone number — a lower identity bar. The honest framing: the biometric apps raise the cost of stolen-photo catfishing, but the most secure dating app is the one where you keep the conversation on-platform, insist on a live video call, and never send money — because those are the three defences no scammer can engineer around, on any app.
Duet, the college-marketed newcomer from WINK TECH LIMITED, publishes one named safety system: enforced selfie face verification with a blue checkmark badge, which it describes as “as reliable as banking security” — its own claim. Beyond that it relies on a 24/7 human moderation team and an anonymous Report & Block flow that includes a financial-solicitation category. Hinge, a Match Group app, pairs its Selfie Verification with a Liveness Check and 3D Face Authentication, deletes facial-geometry data within 24 hours, and states outright that verification “doesn't guarantee the safety of a particular user.” The practical differences: Hinge documents its biometric process in more technical depth, while Duet — despite the campus branding — publishes no student or university-email verification and no background checks. On both apps the trust question is identical: the badge proves a live face, so judge the person by behaviour — live video, a real meeting, and zero money talk.
Start with the badge, then test the behaviour. On Hinge, a verified badge means the account passed a Liveness Check and a 3D facial match to the profile photos; on Bumble, it means a video selfie matched the photos — so an unverified profile that avoids verifying is your first flag. But authenticity of the photos is not authenticity of the person: romance and pig-butchering scams are often run by real people on their own faces. The checks a fake can't pass are behavioural — will they hold a sustained, unscripted live video call; will they meet in person; do they push to leave the app for WhatsApp or Telegram in the first days; and does the conversation ever bend toward money, crypto or a trading platform? Any money ask ends the question — that is never an authentic match. If you're unsure, run the behaviour through our romance-scam checker.
The major dating sites and apps are legitimate, publicly documented companies, and several are publicly listed. That is usually not what someone asking this actually wants to know. The real question sits one level down, because a platform being legitimate tells you nothing about the person messaging you on it. Every safety feature in this index answers whether an account belongs to a real, live human whose face matches the photos. None of them answers whether that human is honest with you. So the legitimacy question splits in two. Is the site real? Almost always yes, and the way to confirm it is to reach the company through its own site rather than through a link somebody sent you. Is this match real? That is settled by behaviour rather than by the platform: a sustained, unscripted live video call, a willingness to meet in person, no early push to move the chat to WhatsApp or Telegram, and no money or investment talk at any point. A legitimate site with a scammer on it is the ordinary case, not the exception.
Tinder publishes more identity checking than most apps in this index, and all of it aims at one problem. Photo Verification takes a video selfie, runs what Tinder calls a “liveness check” to confirm a real person is in front of the camera, and matches that face to the profile photos before granting a “Photo Verified” badge; Tinder's help centre notes the facial geometry captured “may be considered ‘biometric information’ in some places.” ID verification is offered separately, and Face Check™ asks for a biometric scan with a video selfie “in regions where required.” On the messaging side, “Are You Sure?” prompts a sender to reconsider before sending and “Does This Bother You?” prompts the recipient and makes reporting easier, both of which Tinder frames as tools for respectful conversation rather than fraud detection. What none of it covers is money. Tinder runs no background checks: the Garbo partnership launched in March 2022 and ended in August 2023, and no replacement has been announced since. The honest evaluation, then, is that Tinder has largely solved stolen-photo catfishing and has not attempted the harder problem, because a romance or pig-butchering scammer using their own real face passes every one of these checks. Judge a Tinder match the way you would judge any other: live video, a real meeting, and zero tolerance for money talk.