Omi says 60% of its users are face-verified. The scam column kept filling
An optional facial check with wide take-up, and eighty-five reported cases on the same app. Adoption rates are the wrong number to publish on their own.
Two numbers landed together this week, and only one of them was meant as a boast.
Global Dating Insights reported on 25 August that around 60% of users on Omi have completed facial verification, citing the app’s parent, MatchUp UK Ltd. The same report sets that figure against police data carried by Malaysia’s New Straits Times: 462 Malaysians losing RM32.7 million to scams originating on online dating platforms over the four years to June 2026, more than 70% of them aged 21 to 40. Alongside that, 85 reported cases and RM8.43 million are attributed to Omi.
Neither number has been independently audited, and the second is a count of what was reported to police rather than of what happened.
An optional check, by design
Per the report, verification on Omi is optional — a choice the company attributes to legal and privacy constraints in the markets where it operates. Users who skip it get no badge and, the report says, may face restrictions on some features and be classified as higher-risk.
That is a defensible structure, and the arithmetic still bites. Sixty per cent completion means roughly four in ten accounts a user meets have not done it, and the accounts with a reason to avoid a camera are not distributed evenly across that remainder.
The part a face check cannot reach
Oskar Levin, chief information security officer at MatchUp UK Ltd, is reported as saying human-operated scam accounts remain particularly difficult to identify, because the people running them can spend days or weeks building a relationship before any request for money appears.
That is the honest limit of the technology. A liveness check establishes that a live person matches the profile photos. It does not establish who they are, and it has nothing to say about intent. Anyone prepared to spend a fortnight on a mark will clear it on day one and be no less dangerous on day fourteen.
The scale argues the same way. The article cites UN findings putting scam losses in Asia at up to $114 billion, triple the 2023 figure. A badge is a filter applied to that flow, not a wall across it.
We have drawn this distinction before — photo verification became the category standard while identity verification did not, and the two are still marketed as though they were one product. Adoption percentages belong in that pattern. They measure how many people completed a process, not how well the process works.
What would settle the question is data no operator in this category publishes: how quickly reports get actioned, how often a verified account is later removed for fraud, and how many appeals succeed. Until then, treat the badge as one weak signal among several. This is an adults-only category, 18+, and its oldest rule has not moved: a request for money is the tell, however long the conversation has run.