Build the Truth Once

TL;DR

• Vietnam's Ministry of Culture, Sports and Tourism has proposed a “1+1+34+N” governance model for tourism data — one ministry setting standards, one connecting national security databases, 34 provinces maintaining local data, and N other ministries, industries and businesses consuming it.

• This is a governance architecture, not a chatbot announcement — and that is precisely why it matters more. It answers who owns which layer of truth before anyone builds an interface on top.

• The commercial stakes are real: Vietnam took 21 million international visitors in 2025, is targeting 25 million in 2026, and is watching international search demand for the country rise 20 percent, with high-end demand up 38 percent.

• For DOT, the right question isn't “should we build a tourism AI.” It's “who owns which layer of the truth, and how does it connect.”

Bottom line: DOT needs a data constitution before it needs a chatbot — and Vietnam has just published a working draft of what that looks like.

The Signal

Yesterday, Vietnam's Ministry of Culture, Sports and Tourism put a specific, structural proposal on the table. Deputy Minister Hồ An Phong told a national tourism conference that AI and big data are becoming direct drivers of tourism development and called for coordinated adoption across government, businesses, destinations and local communities.

The framing itself is familiar — most tourism ministries in the region have said some version of this in the past two years. What makes Vietnam's announcement different is that it did not stop at the call to action. It proposed an actual governance architecture: a “1+1+34+N” model for who owns, connects, and consumes tourism data.

Vietnam's stated operating principle, presented at the same conference, is worth repeating exactly as framed: shared data, AI support, connected platforms, business partnerships, and tourists at the center. Notice what is first on that list. Not AI. Data.

Why This Matters More Than a Chatbot

It would have been easy for this announcement to be another “we are launching a tourism AI assistant” story. It is not. It is a recognition, at the ministerial level, of a harder and more valuable truth: AI without a reliable destination-data layer beneath it is simply a more articulate way of accessing fragmented information. A well-tuned model can phrase a wrong or outdated answer beautifully. It cannot fix the fact that no one agreed on who owns the answer in the first place.

The 1+1+34+N model solves a governance problem before it touches a technology problem. It assigns exactly one owner for standards and coordination, exactly one owner for the sensitive cross-government data connection, a clearly bounded set of provincial actors responsible for local accuracy, and an open door for everyone else — ministries, airlines, hotels, OTAs — to plug into a layer they did not have to build themselves.

There is real commercial pressure behind the initiative. Vietnam received more than 21 million international visitors in 2025 and is targeting 25 million in 2026. Google search data presented at the same conference showed international search interest in Vietnam rising 20 percent year-on-year, with demand for high-end tourism products up 38 percent. A country chasing that kind of growth curve cannot afford to have its destination information scattered across 50 disconnected places.

What This Means for the Philippine DOT

This is precisely the conversation that needs to happen inside Philippine tourism policy right now — and it is a different conversation from the one usually being had. The question on the table should not be:

“Should we create a tourism AI?”

It should be:

“What authoritative information would any tourism AI need — and who in government is responsible for each part of it?”

Vietnam has effectively answered that question for its own government. The Philippines has not yet asked it formally. A national destination-intelligence layer for the Philippines would need the same kind of explicit ownership map Vietnam has just published — adapted to Philippine government structure rather than copied wholesale:

This is a proposed structure, not a claim that it already exists. The value of laying it out this way is that it forces the actual question DOT needs answered: which agency is the single source of truth for which field, and what connects them. Without that answer, any tourism AI initiative — DOT's own, or a private-sector one built on top of scraped and inconsistent content — inherits the same fragmentation problem Pertlink has flagged in earlier Viewpoint papers.

The Chain DOT Should Be Designing For

Whatever the final governance model looks like, the data chain it needs to support is already visible from the demand side. A Philippine national destination-intelligence layer could connect:

DOT → LGUs → Hotels → Airlines → Airports → Attractions → Restaurants → Transport → Events → Weather → Safety → Payments → Booking.

The interface on top of that chain does not need to be decided now, nor does it need to be exclusive. It could be ChatGPT. Gemini. A DOT-branded application. An airport terminal, in the way Shanghai has just deployed at street level. A hotel concierge desk. An airline's own app. An international travel agent is querying an API instead of calling a regional tourism office. The interface is replaceable. The data chain underneath it is not — and it only needs to be built once.

Three Questions for DOT

1.  Ownership — which office within DOT, or which inter-agency body, plays the role Vietnam has assigned to its Ministry of Culture, Sports and Tourism — standards, schema and coordination?

2.  Connection — which existing government system plays the role of Vietnam's Ministry of Public Security link — the one authoritative bridge into adjacent national databases (immigration, safety, weather, transport)?

3.  Participation — what is the mechanism by which regions, LGUs, and private industry are required — or incentivized — to keep their layer of the data current, the way Vietnam's 34 provinces are expected to?

The Pertlink Bottom Line

Build the truth once. Allow many interfaces to consume it. That is infrastructure.

Vietnam's announcement will likely be reported, in most places it is picked up, as an AI story. It is not one. It is a data-governance story wearing an AI story's clothing — and it is a more useful template for the Philippines than any single AI product launch could be. The competitive question for Philippine tourism in the next eighteen months is not whether DOT ships a chatbot before or after its ASEAN neighbors. It is whether DOT has, by then, answered the ownership question that Vietnam has just answered for itself.

Pertlink's position, unchanged from Part One, is that the interface is the easy part. What Vietnam has demonstrated this week is that the governance model for the data underneath it is not a technical afterthought — it is the actual policy decision, and it belongs squarely on DOT's desk.

The intelligence may be artificial. But the experience is human.

Made with the help of various AI tools – but always with a HITL