Omi's 60% Verification Rate: Safety Measure or Security Theatre?

Omi's 60% facial verification completion rate would be something to celebrate—if the company hadn't also disclosed that 85 cases on its platform resulted in RM8.43M in scam losses. The dissonance between those two numbers crystallises the central problem with voluntary verification: it creates the appearance of safety whilst leaving the door wide open for precisely the sort of fraud it's meant to prevent.

The timing couldn't be sharper. Malaysia ruled out mandatory ID checks for dating platforms just days ago, positioning voluntary measures as the industry's preferred path forward. Omi's parent company MatchUp UK Ltd now offers the highest-profile test case for whether self-regulation can actually work.

Chief information security officer Oskar Levin told the New Straits Times that whilst 60% of users have completed the optional facial scan, those who don't face feature restrictions and higher-risk classifications. Users who pass receive a verification badge.

Create a free account

Unlock unlimited access and get the weekly briefing delivered to your inbox.

Register free

No spam. No password. We'll send a one-time link to confirm your email.

What's missing from that announcement is rather more telling than what's included. The company hasn't disclosed whether that 60% figure represents active users or total registered accounts. The timeframe for reaching this rate is also absent. And there's no data on whether the 85 fraud cases reported to Malaysian police involved verified or unverified accounts—which would be the single most important piece of information for assessing whether the system actually works.

The Verification Paradox

A 60% completion rate sounds like industry-leading performance until you remember that the remaining 40% of users—millions of people across Omi's Asian markets—can still message, match, and build the trust relationships that precede fraud.

If voluntary verification creates a two-tier system where legitimate users feel compelled to verify whilst scammers simply migrate to the unverified pool, the entire exercise becomes security theatre. Omi's own loss figures suggest that's already happening.

Feature Restrictions as Coercive Design

Levin characterises facial verification as optional due to 'legal and privacy considerations', but the operational reality suggests something rather different. Restricting features for unverified users whilst flagging them as higher-risk effectively creates a coercive design pattern that makes verification mandatory in all but name. Privacy-conscious users face a choice: hand over biometric data or accept reduced functionality and potential stigmatisation.

That design choice matters because it shifts the framing from legal mandate to platform architecture. Malaysia's decision to rule out mandatory ID checks becomes less significant when dating operators can achieve the same outcome through product design. The government avoids the political cost of mandating biometric surveillance whilst platforms get the verification rates they need to claim they're serious about safety.

Whether this constitutes genuine safety infrastructure or simply liability protection remains an open question. According to police data reported by the New Straits Times, 462 Malaysians lost a combined RM32.7M to romance scams over the four years to June 2026, with more than 70% of victims aged 21 to 40. The 85 Omi cases alone account for roughly a quarter of that total—a remarkably high concentration for a single platform.

AI Escalation Outpaces Voluntary Measures

Levin's comments about scammer tactics underscore why voluntary verification may be structurally insufficient. According to the CISO, human-operated scam accounts remain particularly difficult to identify because fraudsters can spend days or weeks building relationships before attempting to extract money. Automated bots are easier to detect due to predictable behaviour patterns, though AI has made these systems more flexible and responsive.

That assessment tracks with broader regional trends. UN figures reported by the Philippine Daily Inquirer show scam losses across Asia reached $114B in 2025—triple the 2023 figure. Romance scams represent only a portion of that total, but the trajectory suggests the threat is evolving faster than platforms can respond. AI-generated profiles that pass facial verification, deepfake video calls, and sophisticated social engineering all render traditional verification methods less effective.

The fundamental problem is that facial verification confirms you are who you claim to be—but it doesn't confirm you're not a scammer. The verification badge becomes a trust signal that victims interpret as vetting when it's merely authentication.

That gap between what verification promises and what it delivers may be the industry's most significant unaddressed vulnerability.

The Southeast Asian Context

The Southeast Asian context adds further complexity. Many markets in the region lack the robust identity infrastructure that makes verification meaningful in Western markets. Stolen or fabricated documents can pass verification checks, particularly when platforms rely on automated systems rather than manual review. Omi hasn't disclosed what verification actually entails beyond 'facial scanning', leaving open questions about document checks, liveness detection, or database cross-referencing.

For operators watching this unfold, Omi's experience offers a preview of what voluntary verification looks like at scale. A 60% completion rate is likely near the ceiling for optional systems—the truly privacy-conscious or technically challenged won't verify regardless of feature restrictions, whilst scammers will adapt their tactics to exploit whatever gaps remain. The question isn't whether voluntary measures are better than nothing. The question is whether they're sufficient to meaningfully reduce fraud, or whether they simply create the appearance of action whilst losses continue to mount.

Malaysian regulators clearly believe the industry deserves the chance to prove self-regulation can work. Omi's RM8.43M in reported losses suggests that proof may be harder to establish than the 60% headline figure implies.

D