Before a profile goes live
Generated faces, stolen photos and duplicate accounts are caught at sign-up rather than after the first report. A held profile costs one person a delay; a live fake profile costs several people money.
WHITE LABEL · PROTECTION
Romance scam protection, in the message path
Romance fraud is not opportunistic any more. It is run as a business, with scripts, shift patterns, managed account inventory and a pipeline for cashing out. Protection designed for opportunistic bad behaviour does not touch it.
While the conversation is happening
Chat on D8N doubles as a trust sensor. Message risk is assessed in the path rather than in a nightly batch, which is the only timing that helps: a warning delivered the next morning arrives after the transfer. A money request from a recent match raises an elevated risk state with reasons attached, and the recipient sees a warning in the conversation where it is relevant.
Linked accounts fall together
Scripted accounts running the same opener across a fleet are linked by behaviour and removed together, rather than one at a time as reports trickle in. Taking down one account out of forty is not enforcement; it is a rounding error.
And they stay out
D8N Guard stops confirmed abusers from returning under a new account — and the block holds across every brand on the platform. That shared enforcement is the reason a small brand on D8N gets protection its own traffic volume could never justify building.
Payment fraud, before the chargeback
Risky cards on minutes-old accounts are declined at the upgrade rather than reconciled later. Chargebacks are expensive twice: once in money, and once in the processor relationship you need in that market.
QUESTIONS
Straight answers
If the answer you need isn’t here, ask us directly — we will tell you plainly when D8N is not the right fit.
How do you detect a romance scam without reading everyone's messages?
Assessment is automated and scoped to risk signals, and it returns a state with reasons rather than a transcript to a human. Operators see message evidence when a case is opened — that is, when a report or a strong signal has already justified it — not as a matter of routine.
Do members see a warning, or do you just block?
Both exist, and they escalate. Warning, then challenge — verify again — then restriction, then investigation and enforcement. Blocking on a weak signal trains members to ignore the product; warning on a strong one leaves them exposed. The state decides which.
Does a ban on one brand apply to the others?
Yes, for confirmed abuse. D8N Guard blocks new accounts from an actor banned anywhere on the platform, and the pattern is learned network-wide rather than per brand.
Is a D8N-powered app really my brand?
Yes. Your name, domains, theme, copy, pricing, app store listings and member relationships are yours. The only D8N marks members see are the trust marks — Protected by D8N and D8N Verified — which exist because a visible mark is what makes verification worth passing.
Where do I start?
Tell us the market, the audience and whether you are launching a new brand or adding D8N to an app you already run. Those three answers decide whether white label, the API, or a custom build is the right shape, and we will say plainly if it is none of them.