You do not need her ID to know she is real. You need a reverse image search, a live video call with a couple of unannounced moves, and one rule about age that ends things the moment it’s unclear, in that order, cheapest first.
Most men run these checks backward: asking for personal documents before they’ve done a free image search, or treating a platform verification badge as proof. It isn’t. A selfie-liveness badge proves a face matched a video at signup, not that she is who her profile claims, and it never proves age.
Why you never ask for her ID
Asking for a government ID is worse than useless. If you collect it and it ever leaks, you’re the reason, and that’s a liability sitting on your hard drive with no upside. It’s also a coercion signal: demanding her driver’s license photographed over text isn’t screening for safety, and a woman with any experience in this reads the ask as either naive or the setup for something worse. In several states, collecting a face image alongside ID runs into biometric-privacy law on top of that.
The fix: ask her to complete the platform’s own ID or photo verification and check the badge, rather than collecting anything yourself. That tells you a document matched a face, without the document ever touching your phone.
The order of checks, cheapest first
Run these in sequence, stopping the moment one fails. Most fake profiles die at step one or two.
-
Reverse image search the photos (minutes, free). Crop each photo to the face and run it through Google Lens and Yandex Images; Yandex is meaningfully better at matching the same face across different photos. A hit under a different name, or on a stock or modeling site: stop. No hits proves nothing either way; private people and AI-generated faces both return nothing.
-
Check the AI tells. A profile that’s one studio-lit session, with no candid shots, no other people with sharp faces, no identifiable public place, and jewelry that shifts between photos reads as generated or scraped. Hands and garbled background text are worth a glance, but they’re weak signals: production-quality fakes have mostly fixed the old tells, and ordinary people spot AI faces barely better than a coin flip. Your gut triggers the other checks; it isn’t a verdict.
-
Username and social footprint. Run the handle through WhatsMyName. On any linked Instagram or TikTok, check account age (the “about this account” panel shows join date), whether other people’s tagged photos and comments exist, and whether face, location, and apparent age hold consistent across years of posts. A weeks-old account with a fully built “life” already on it is the classic scam signature.
-
The conversation test. Ask something tied to what she just told you, something locally specific, or something time-sensitive (“what’s the weather where you are right now?”) and check it. Scripted operators, increasingly AI-assisted, answer generically or steer back to logistics. Watch timing too: instant replies to anything emotional but slow, generic answers to a factual question is a tell for a multi-chat operator, human or automated. An early push to move off-platform before more than a few messages is a flag, not a disqualifier alone.
-
The video call. The real test, and non-negotiable before anything else happens. Covered below.
The video call: ask for things she can’t have scripted
A recorded video or a call where nothing happens proves almost nothing. What matters is asking for small, unannounced physical actions, live:
- Turn to a full side profile and hold it.
- Wave a hand slowly across her face, or hold fingers over her nose and mouth for a second.
- Stand up, step back from the camera, or walk to another room.
- Turn a room light off and on, or point the phone toward a window.
- Hold an object with text on it up next to her face.
- Read back a short phrase you just picked, out loud.
These target the weak points of real-time face-swap tools: occlusion, profile angles, and lighting that doesn’t track across the swapped face. None of this is foolproof, and a scammer only needs the call long enough to record you, so treat it as one input, not a certificate. A refusal, for any stated reason, is the strongest signal there is. Don’t argue with it. Stop.
The words on a profile that mean walk away
Some phrasing tells you what you’re looking at before you’ve spent a minute on anything else:
- A fixed price for a specific act, or “PPM” framed as an hourly rate rather than a meet: that’s not sugar dating, that’s a commercial-sex negotiation, with the legal exposure attached.
- A push to skip the meet-and-greet for a paid meeting, or urgency about “not wasting time.”
- Any request for money, a deposit, a “processing” or “verification” fee, or gas money before you’ve met in person, however small. This is the single most reliable tell across every version of the fake-profile playbook: the ask always arrives before the meeting.
- Vague, generic warmth that never answers a specific question, paired with steady pressure to move to a different app.
A real conversation looks different: she engages with what you wrote, asks her own questions, and talks logistics, including money, without either of you producing proof of anything. Ordinary back-and-forth about expectations is normal; a demand for cash before a first meeting is not.
The age rule: any ambiguity ends it
This comes before everything else above. If anything about her stated age is unclear, inconsistent, or walked back mid-conversation, stop immediately: before another message, before a call, before a meeting. Don’t try to resolve it yourself by asking for ID; end the conversation and move on.
There’s no step after that one. Age ambiguity isn’t a caution flag to weigh against everything else you’ve verified. It’s a stop condition, and it’s the one check that costs nothing and takes no time to run: read what she actually told you.
Between the video call and this one rule, you’ll filter out nearly everything worth filtering. The free download on this page has the full checklist in order, so you can run it in a few minutes before agreeing to meet.
Sources: FBI IC3 2025 Internet Crime Report; FTC consumer guidance on romance scams; Seeking platform safety and verification pages; PimEyes and FaceCheck.ID product documentation; academic study on human detection of AI-generated faces (Royal Society Open Science, 2025); reporting on real-time deepfake video-call scams (Hong Kong police, 2024; France24, 2025). General education, not legal advice.