The Trust Lag: Lessons from Vietnam's Digital ID Rollout

Infrastructure and trust move on different clocks. You can build a national digital identity system in a few years—or in less than one if you leverage global open-source platforms (see MOSIP adopters including the Philippines, Ethiopia, Madagascar, and Togo). Getting a population to actually believe in it takes longer and much more than a press release. 

Vietnam is a live case study in that gap. VNeID, the country's national digital identity app, is being built out as a ‘super app’ under a government roadmap running through 2045. A new Personal Data Protection Law took effect in January 2026. Biometric verification became mandatory for banking the same month. This is one of the fastest-moving digital public infrastructure programmes anywhere, and it's genuinely impressive in scale. It's also running into the same problem every fast DPI rollout eventually hits: the tech can outpace the trust.

Three moments this year show what that looks like in practice.

The app that crashed on good news. In August 2025, the government announced a small cash payment to citizens. VNeID buckled under demand almost immediately. It wasn't the first time either, a residency-update deadline had caused a similar outage two months earlier. The pattern is worth sitting with: the moments the system most needed to work, because people were actually paying attention, were the moments it failed. A comms win turned into a capacity problem, on the same channel meant to build confidence.

One number, two (three) stories. Vietnam's central bank deactivated more than 86 million bank accounts for lacking biometric verification. Official coverage called it a data cleanse: mostly inactive or unverified accounts, a fraud-prevention measure. Some international outlets ran with “digital surveillance clampdown”. Independent fact-checks landed somewhere in between, confirming the number but not the scale of harm claimed on either side. The problem is a vacuum that size, left under-addressed by the primary source, gets filled by whoever tells the most vivid story first.

The quiet reversal. In May 2026, police stopped collecting DNA and voice samples for the ID database after public pushback. That's genuinely good evidence that feedback loops exist and that policy moves in response to them. It barely registered as a trust-building moment, because nobody framed it as one.

Underneath all three is a more concrete problem: who actually gets left behind when verification becomes mandatory. Viettel reported roughly 5 million unverified subscribers ahead of a June 2026 deadline, and by its own account, most weren't holdouts. They were elderly, rural, remote, or too unwell to travel and verify in person. It's a specific, countable population, and exactly who a comms strategy needs to plan for from the beginning, not scrambling to reach afterwards.

None of this is unique to Vietnam. The World Bank's own digital public infrastructure programme names trust, inclusion, and security-by-design as core to the model. Every country moving fast on digital ID or payments infrastructure hits some version of this tension, because the tension is structural. Vietnam is just a current, well-documented example of a pattern showing up everywhere DPI gets built at speed.

Which is the point. Communications isn't the layer you add once the infrastructure is live, to explain it to people after the fact. It's the mechanism that either builds trust ahead of the friction, or has to repair it once the friction has already made the news. Three things separate the two:

Publish the number before someone else does the presumptuous maths. If 86 million accounts are being deactivated, say so clearly, with the breakdown, before headlines start filling in the blanks.

Build the exclusion story into the rollout plan, not the apology. If a mandatory deadline will disproportionately hit elderly or rural users, that's a problem to solve in the plan, not a line to manage after.

Make responsiveness visible. The DNA and voice data reversal should have been one of the most trust-building moments of the year. It wasn't, because it was never framed as proof that feedback works. That's a genuinely fixable gap, and a cheap one to close.

As more governments and multilateral programmes build this kind of infrastructure, the lesson worth taking from Vietnam is not about the technology stack. It's the discipline around it. The countries that get this right won't be the ones that move fastest. They'll be the ones that treat trust as something built on purpose, with the right communication, at the same pace as the infrastructure itself.