top of page
40051abd5a76713af8f015988fc6780e-blue-phone-icon-with-a-wave-on-it.webp
Logo-Zalo-Arc.webp
Facebook_Messenger_logo_2020.svg.webp
whatsapp

Vietnam’s Data Security Law: from digital sovereignty to new requirements for businesses

6 days ago
9 min read

Vietnam’s draft Data Security Law is shaping a broader approach to cybersecurity in which data is not only an asset to be used, but an asset that must be protected throughout its lifecycle. For businesses, the impact goes beyond legal compliance. Organizations may need to rethink how they classify data, control access, manage cross-border data flows, secure cloud and AI environments, detect abnormal activity, and prepare for incidents that affect critical information.


As digital transformation accelerates, business data increasingly moves across multiple environments: internal infrastructure, cloud platforms, SaaS applications, mobile systems, APIs, AI services, and third-party platforms. Customer information may enter through a website, move into a CRM, be processed in the cloud, downloaded to endpoints, shared with partners, and eventually appear in multiple backup locations.

This changes the nature of the security problem. A system may continue operating normally even while the data inside it has been copied, altered, shared with the wrong party, or used outside the organization’s control.

vietnam-data-security-law
Vietnam’s Data Security Law: from digital sovereignty to new requirements for businesses

The draft Data Security Law is therefore significant because it places the data itself at the center of security governance.

For businesses, the more important questions are no longer simply how many security tools have been deployed, but:

Where is critical data located, who can access it, which systems does it move through, and can the organization detect when control over that data has been compromised?

1. How Is the Data Security Law changing the way businesses think about data?

One of the most important aspects of the draft law is not simply the addition of another regulatory layer. It is the shift toward treating data itself as a core security asset.

Traditional security programs often begin with infrastructure: which servers need firewalls, which endpoints require EDR, which accounts need MFA, or which networks should be segmented.

An organization may already operate multiple cybersecurity technologies and still lack effective control over its data if it does not know how many endpoints contain a sensitive file, which SaaS platforms have synchronized copies, or which external providers retain backups.

In other words, security can no longer stop at the perimeter of the system holding the data. Security needs to follow the data as it is created, used, shared, and retained.

IPSIP Vietnam previously examined the policy background and why Vietnam is moving toward a dedicated data security framework in Draft Data Security Law: What Should Businesses Prepare For?. This article focuses more specifically on the operational, governance, and cybersecurity implications for businesses.

2. How does the Data Security Law differ from existing legal frameworks?

Businesses in Vietnam are already operating under multiple legal frameworks that address data from different perspectives.

At a high level, the distinction can be understood as follows:

Legal framework

Primary focus

Data Law

Governance, use, sharing, and organization of data

Personal Data Protection Law

Rights, obligations, and conditions for processing personal data

Cybersecurity Law

Security of information systems, network infrastructure, and activities in cyberspace

Draft Data Security Law

Security of the data itself based on risk and throughout its lifecycle

This table is intended only as a high-level comparison and should not replace a detailed review of the scope and obligations under each legal framework.

In practice, the same dataset may fall under several frameworks at once. Customer information may qualify as personal data, reside within systems subject to cybersecurity requirements, and still require broader governance and protection throughout its lifecycle.

3. What Does “Data Sovereignty” Mean in Practice?

In a digital environment, data sovereignty is closely connected to control: who can access the data, which organizations are processing it, where it is transferred, which systems hold copies, and whether the business can revoke or restrict access when necessary.

This becomes particularly important when organizations use:

  • international cloud platforms;

  • CRM and SaaS services;

  • AI platforms;

  • managed IT or managed security providers;

  • outsourced software development;

  • systems operated by teams across multiple countries.

A business can use cloud services while maintaining strong governance if it clearly understands its data flows, storage regions, access rights, key-management practices, retention policies, and all parties capable of processing the data.

The reverse is also true. Data stored on an on-premises server is not automatically secure if privileged accounts are broadly shared, encryption is absent, and access is not properly logged.

4. How do the five risk levels change the way businesses protect data?

One of the notable elements in the draft is the use of five data security risk levels.

The underlying principle is that protection should correspond to the potential level of damage if the data is compromised.

5-risk-levels
How do the five risk levels change the way businesses protect data?

This approach can also help organizations allocate resources more effectively. Without data classification, businesses may fall into one of two extremes: applying unnecessarily strong controls to all information at excessive cost, or using a baseline that is too weak for critical data.

It is also important not to treat the draft law’s data-risk levels as identical to existing information-system classification models.

Businesses can review IPSIP Vietnam’s article on five cybersecurity levels businesses should understand in 2026 for additional context on risk-based protection, while keeping in mind that systems and data are different assessment objects.

5. Data lifecycle protection may be the biggest operational change for businesses

Data does not remain in one place. A typical business data lifecycle may include:

Creation → Collection → Storage → Processing → Transmission → Sharing → Backup → Archiving → Destruction

This is why protecting only the primary database is no longer enough. The main database may remain secure while additional copies of the same information move beyond the original control boundary.

Each stage of the lifecycle also creates different risks:

  • During collection, businesses need to determine whether they are gathering more information than necessary.

  • During storage, the focus may shift to encryption, access control, and backup.

  • During sharing, risks involve recipients, permissions, access duration, and download rights.

  • During backup, organizations need to know who can read, restore, or delete copies.

  • During destruction, data should be handled according to retention policies rather than simply disappearing from a user interface.

A mature data security strategy therefore needs to answer: Where is the data? Who is using it? Where is it being transferred? How many copies exist? What happens when the data is no longer needed?

6. Cross-border data will become a larger governance challenge

Cross-border-data
Cross-border data will become a larger governance challenge

Cloud and SaaS make cross-border data flows increasingly difficult to see unless organizations deliberately map them.

A CRM may store information in another region. Backup systems may replicate data to overseas data centers. A vendor’s support team may access systems from several countries. An application using an external AI API may send prompts and related information outside the organization’s environment.

For that reason, cross-border data governance should not begin only with the question of whether data can be transferred.

Businesses first need to understand:

What data → Where it goes → Who processes it → For what purpose → Who can access it → How long it is retained → Which controls protect it

7. How is AI changing the data security problem?

AI is creating an entirely new layer of data flow.

Employees may upload internal documents to chatbots for summarization. Retrieval-Augmented Generation systems may connect directly to corporate knowledge repositories. AI agents may receive permission to read email, access CRM systems, work with cloud drives, or take actions on business systems.

The key point is that AI access should not be treated as less sensitive than human access. If an AI agent can read an entire repository or act using privileged credentials, the potential impact of misuse, misconfiguration, or compromise can be significant.

8. Which businesses should pay the closest attention?

High data-security risk is not limited to banks or government agencies. Organizations that should pay particular attention include:

  • banks, fintech companies, and financial-service providers;

  • healthcare organizations;

  • telecommunications companies;

  • e-commerce businesses;

  • SaaS and cloud providers;

  • logistics companies;

  • foreign-invested enterprises;

  • businesses holding large volumes of customer data;

  • operators of critical infrastructure;

  • companies processing data on behalf of other organizations.

Data volume and data value are not the same thing.

A dataset containing only a few hundred privileged credentials, financial records, or trade secrets may create significantly more risk than millions of public records. Businesses should therefore avoid using record count as the only indicator of data-security risk.

9. What should businesses prepare for in 2027?

For many organizations, 2027 is likely to mark a transition from simply monitoring regulatory developments to building practical data-governance and security capabilities.

Rather than waiting until every final requirement is in place, businesses can prioritize foundations that remain useful for both compliance and cybersecurity regardless of how individual provisions are ultimately implemented.

  • First, organizations should identify where critical data actually resides. This means looking beyond the primary database to include files on endpoints, SaaS platforms, cloud storage, backups, vendor systems, and additional copies created during day-to-day operations. Without a sufficiently clear data inventory, subsequent risk classification and security controls are difficult to apply accurately.

  • The next step is to classify data according to its importance and the potential impact of compromise. As the draft framework moves toward risk-based management, businesses should avoid using the same security baseline for every category of information. Public information, internal data, financial records, privileged credentials, and trade secrets require different levels of protection.

  • Organizations should also review access rights and cross-border data flows. Cloud services, SaaS, AI, and third-party environments mean information may be processed across multiple locations without being immediately visible to internal teams. Businesses need to understand what data leaves their environment, who processes it, who can access it, and which controls apply at each stage.

  • From a technical perspective, 2027 is also an appropriate time to reassess core protection layers such as access control, encryption, backup, logging, monitoring, and incident response. The goal is not to deploy as many technologies as possible. It is to determine whether critical data is being protected in proportion to its actual risk.

  • Finally, organizations need to close the gap between “having a policy” and “having controls that actually work.” A documented access policy does not necessarily mean permissions are properly controlled. A backup requirement does not guarantee that information can be recovered. Monitoring technology does not automatically mean abnormal access will be detected.

The objective for 2027 should therefore go beyond simply complying with a new law. Businesses should work toward a state where they can:

Identify what data needs protection → Know where it resides → Understand who can access it → Know which controls are protecting it → Detect and respond when control is compromised.

That is the foundation for making compliance and data security part of the same operating model rather than two separate programs.

10. Which security measures should businesses prioritize?

No single technology can solve the entire data-security problem. A more effective model combines:

Classification → Access Control → Encryption → Monitoring → Backup → Incident Response

Two layers are particularly complementary: protecting the data itself and detecting attempts to compromise it.

ipsip-viet-nam-cybersecurity-solutions
IPISP Vietnam cybersecurity solutions

10.1. Data encryption: protecting the asset directly

Encryption operates directly on the data. If a file, database, or storage environment is accessed without authorization, properly encrypted information creates another barrier preventing someone without the appropriate key from immediately reading the content.

Encryption is particularly relevant for:

  • sensitive information stored on servers;

  • cloud data;

  • backups;

  • information exchanged between systems;

  • data shared across multiple environments.

For critical information stored across servers, cloud platforms, or distributed environments, organizations can review IPSIP Vietnam’s enterprise data encryption solution as one possible technical layer within a broader data protection strategy.

10.2. SOC 24/7: when data protection also requires detection and response

Encryption can reduce the risk that unauthorized parties are able to read data, but organizations still need to know when someone is attempting to reach or extract that data.

This is where a Security Operations Center can add another layer of protection.

A SOC does not protect data in the same way encryption does. Its role is to monitor signals from identity systems, endpoints, networks, cloud platforms, and security tools to identify behavior that may indicate that data is under threat.

For organizations that cannot maintain continuous in-house security monitoring but still require ongoing detection, investigation, and escalation capabilities, IPSIP Vietnam’s SOC 24/7 service can be considered as one operating model for strengthening monitoring and incident detection.

Vietnam’s Data Security Law reflects an important shift in cybersecurity thinking: protecting systems remains necessary, but it is no longer sufficient if businesses cannot control the data that exists within and moves across those systems.

This becomes increasingly important as information moves across cloud platforms, SaaS services, APIs, endpoints, AI systems, and third-party environments before completing its lifecycle.

Preparing for the Data Security Law should therefore not be viewed only as a compliance project. It is also an opportunity for organizations to reassess how they identify, govern, protect, monitor, and respond to data-related risk in an increasingly cloud- and AI-driven economy.

References

follow ipsip vietnam.png
bottom of page