Ho Chi Minh City’s digital transformation is moving from policy to infrastructure: as of September 2026, the city has collected 25 of 39 required data categories (64%) under Scheme 06, standardized 31,768 locations with relatively complete GPS coordinates and administrative unit data, and connected its city data warehouse to Vietnam’s National Data Center in a testing environment, according to VnEconomy (October 11, 2026). Those numbers frame what is actually a harder problem, bringing fragmented government data into a unified system capable of powering AI-driven public services.
Why is data fragmentation still the core bottleneck in HCM City’s digital transformation?
Different city agencies hold separate datasets that don’t connect, and for large language models (LLMs, AI models trained on massive datasets to understand and generate language), the quality of input data determines the quality of output. Nguyen Thanh Hoa, Deputy Director of HCM City’s Digital Transformation Center, told VnEconomy that data must meet six conditions before it can reliably power AI models or data mining tools: it must be correct, complete, clean, current, unified, and shareable.
The city is addressing this through its IOC (Integrated Operations Center, a unified hub that monitors and processes live data streams from across city departments and districts) at 6 Tu Xuong, designed to connect data from wards, communes, special zones, and agencies into a single real-time pipeline. A 100-day intensive campaign running from July 10 to November 30, 2026 is targeting the remaining digital transformation bottlenecks. Nguyen told VnEconomy the center will provide a fuller assessment after the campaign closes.
The legal foundation has also strengthened. Politburo Resolution 57-NQ/TW (December 22, 2024) set the national framework for science, technology, innovation, and digital transformation. Vietnam’s Data Law now defines which agencies own and create specific datasets, giving coordination a clearer legal basis. The National Data Architecture Framework specifies which applications ministries will standardize for shared nationwide use, so local governments won’t need to duplicate that investment.
What has the Digital Citizen app achieved, and where are the gaps?
HCM City’s “Cong dan so” (Digital Citizen) app has surpassed 1.2 million downloads and logged more than 7 million interactions, per VnEconomy. Current integrations include:
- A public safety SOS miniapp (a lightweight application embedded within a larger app) from the city Police Department
- Education data and student academic records
- Healthcare data
- Trade union member records: over 1.8 million profiles integrated out of an estimated 2.2 million union members citywide
Nguyen noted that for a city of HCM City’s population scale, 1.2 million installs still leaves significant room to grow. The key factor driving sustained usage, he told VnEconomy, is utility: technology must solve real problems to bring users back.
The city’s longer-term direction points toward a super-app model where residents access multiple services through a single platform. Nguyen suggested that VNeID, Vietnam’s national digital identity system, could serve as that central layer, built on unified authentication infrastructure and capable of hosting services across sectors. He cited models from cities in China, the United Kingdom, Australia, and the United States, where digital apps have integrated a range of needs, from watching television and reading news to ticket information, discount programs, and essential services.
How is HCM City building data center capacity for AI workloads?
HCM City currently hosts approximately 20 of Vietnam’s 40-plus data centers. Vietnam’s national target is 500 MW of total data center capacity by 2030. Investment costs run at roughly $6 to $8 million per MW, according to figures Nguyen cited to VnEconomy, so facilities of several hundred MW require a very large total investment. The city has identified potential sites including the High-Tech Zone in Thu Duc, Tan Thuan, and Cu Chi’s Tan Phu Trung Industrial Zone.
Two supply-side constraints limit faster development: stable power supply, and access to high-performance computing chips (specialized processors, primarily GPUs, required to run AI workloads at scale), which remain subject to the export licensing policies of supplying countries. Nguyen told VnEconomy that resolving chip access for high-performance computing centers remains an active challenge.
On August 28, 2026, Prime Minister Decision 1671/QD-TTg established a National AI Strategy through 2030 with a vision to 2045. Target sectors for AI application include agriculture, logistics, aviation, and transportation.
What role will AI Agents play in HCM City’s next phase?
Nguyen drew a distinction between generative AI (Gen AI, systems that produce text, images, and other content on demand) and agentic AI (AI Agent, systems capable of retaining context across sessions, planning multi-step actions, and executing tasks autonomously). He told VnEconomy the city plans to push AI Agent deployment into specific operational use cases in 2027.
Agentic AI could handle tasks previously requiring human staff, he said, and could enable new business models, including single-person enterprises that run multiple AI agents handling different functions while the human owner focuses on oversight and final decisions.
What are the data privacy and security risks as data consolidates?
Vietnam’s personal data protection framework now rests on the Law on Personal Data Protection and Decree 356/2025/ND-CP, which superseded the earlier Decree 13/2023/ND-CP. Nguyen told VnEconomy that scam call volumes have decreased notably through coordination with Interpol and international partners.
However, cyberattacks remain a concern worth monitoring closely. Attack frequency may decrease, but the severity of individual incidents could increase, he noted. Ransomware attacks (cyberattacks that encrypt an organization’s data and demand payment for its release) could cause large enterprises losses of millions of USD. His guidance: organizations must actively back up data, maintain contingency plans, and prepare specific defenses against new attack methods, rather than waiting for incidents to occur.
Is digital literacy keeping pace with infrastructure?
HCM City launched its “Binh dan hoc vu so” mass digital skills campaign on October 11, 2025. The Digital Transformation Center alone has since run approximately 174 classes reaching around 38,000 participants, per VnEconomy. Nguyen noted a demand shift: residents are moving beyond general digital awareness toward practical, task-specific skills. The focus for the current period, he said, is not just knowing how to use technology but being able to improve and innovate with it.
What this means for AI-search visibility
Fragmented government data doesn’t just slow down city administration. It makes a city’s businesses, services, and locations harder for AI systems to find, verify, and accurately describe to users who ask about them.
Here is the concrete implication. When a user asks an AI assistant about hospitals in a specific HCM City district, or about permit processes, the answer can only be as good as the data the AI learned from or can retrieve. In Hingewise’s view, a unified, standardized city database with accurate location coordinates, service categories, and current status is more likely to show up in a coherent AI response than fragmented records scattered across disconnected agency systems. The standardization of 31,768 location records with GPS coordinates, reported in VnEconomy’s October 2026 interview, is, in Hingewise’s view, an input to entity resolution: the process by which systems link a name, a location, and a set of attributes into a consistent profile. Without that resolution, AI systems tend to guess, hedge, or omit.
In Hingewise’s view, Vietnam’s most consequential AI readiness gap right now is data completeness at the entity level. With 25 of 39 required data categories collected under Scheme 06, a meaningful share is still missing. For AI systems that index and cite Vietnam’s digital economy, those gaps are not neutral: they translate into blind spots where businesses and services exist in reality but are underrepresented or misrepresented in AI-generated responses. Brands and services in HCM City that are well-represented in structured, machine-readable datasets, whether through Digital Citizen app integrations, VNeID connections, or standardized location registries, will likely achieve higher Share of Model (the rate at which AI systems accurately mention and describe them relative to competitors) than those that exist only in poorly structured or inaccessible sources.
The $6 to $8 million per MW data center investment figure also signals something beyond infrastructure planning. As Vietnam’s compute capacity grows, so does the local AI inference capacity that determines how Vietnamese users experience AI-powered services. That makes the chip access constraint worth tracking carefully. Compute capacity ultimately determines whether Vietnam’s well-structured data translates into capable local AI systems or whether it mostly feeds AI platforms built and hosted elsewhere.
What to follow: whether HCM City’s 100-day campaign, closing November 30, 2026, moves its data warehouse connection with the National Data Center from a testing link to a production-grade integration, and whether VNeID’s development trajectory advances toward the kind of structured, machine-readable service layer that AI systems can draw from as a reliable primary source.
Source: VnEconomy, October 11, 2026. Interview with Nguyen Thanh Hoa, Deputy Director, HCM City Digital Transformation Center. Read the original (Vietnamese).
This article was drafted with AI assistance from the sources cited above and checked against them before publishing. The "What this means for AI-search visibility" section is Hingewise's editorial analysis, built on the frameworks and positions we use in our GEO work. How we work: Editorial policy.