Vietnam’s Personal Data Protection Law Has Been in Force for Six Months. It Is Quietly Rewriting How Digital Lenders Score Borrowers

The Vietnamese Personal Data Protection Law No.91/2025/QH15 and Decree 356/2025/ND-CP came into force on January 1, 2026. Half a year on, the more tangible results have been compliance workstreams and the implementation of privacy notices. The bigger of the two problems is structural and lurks below Vietnam’s blistering lending market.

According to a recent report from Vaynhanh Research called the Vaynhanh Fast Loan Market Report 2026, the overall outstanding balance of accounts with fast credit characteristics reached about VND 1,100 trillion in mid-2026 (around US$42 billion or some 5% of Vietnam’s total system credit: VND 20.03 quadrillion) (Báo Chính phủ, 2026). Much of that growth this last year has relied on evaluating borrowers who have had no traditional credit file in the past five years.

This is a business, data-dependent, and the data rules changed in January.

Why a data law lands on lending specifically

Rather than products, the fast credit market in Vietnam is better defined in terms of process. A working definition of a report refers to small to medium-sized personal credit, relatively less documented, quicker in the decision than secured lending and much easily available.

All those attributes are bought with data. Verification from third-party sources reduces the documentation burden. Automated assessment accelerates decisions. Digital accessibility simply means the entire borrowing process leaves a trail of data. Without the data layer, the product reverts to a slow and shallow version of itself.

This is precisely why a generic data protection law is not just a bit-by-bit compliance issue in this market. It works at the input level, enabling the product to be made.

The regulatory stack is now doubled upTwo of the sandbox’s three solutions are data solutions

The controlled testing mechanism for Vietnam’s banking sector was introduced by Decree 94/2025/ND-CP, issued on July 1, 2015. It deals with three solutions: credit score, data sharing through open API and peer-to-peer lending.

Two of those three things are data solutions. Since the start of 2026, fintech testing credit scoring or open API data sharing under the sandbox from mid-2025 onwards is operating under a personal data regime, which was not envisaged when designing the sandbox.

The English-language coverage of Vietnamese fintech has neglected that intersection, which is the practical question faced by anyone looking at the market, what can be legitimately used for a borrower assessment and on what basis, with what record of consent.

How the report treats peer-to-peer lending

Vay Nhanh Research adopts an intentionally cautious position. It’s worth noting that the scope of the report only includes institutions whose certificates of participation in the Sandox mechanism under Decree No.94/2025/ND-CP have been granted by state agencies. Not all so-called P2P platforms are in that definition, nor is every platform in the national scale scope, and no one should presume being part of such regulated bodies.

This is relevant to anyone reading market size estimates for Vietnamese P2P lending. While they might appear similar at first glance, the regulatory status of a platform and how it describes itself are quite different matters and the report keeps these separate.

The hypothesis this creates for 2027

The report puts forward five hypotheses for testing in 2027, including that alternative data could bring a new class of borrowers into the fold who can be properly assessed. The reasoning is straightforward. Your borrowers may be able to repay, but they are not suitable for that particular product because their file is thin, they don’t have the paperwork, they picked the wrong product or they did not know how to compare products. This is partly seen as an information and matching gap, rather than a pure creditworthiness problem, and we have been focused on addressing exactly this kind of structural issue.

One such tool that will help to close it is alternative data. The data protection regime defines the limits on how that tool can be employed.

The report makes it clear that this is merely a hypothesis, marked [H] in its evidence hierarchy, and not an established observation. It recommends key watch items: The shifts such as underwriting criteria change, what data may legally be used and also credit outcomes among thin-file applicants.

Trust and privacy as competitive assets

The second is a related hypothesis. The report proposes that the elements of trust, privacy and transparency may well end up being more competitive assets – more challenging to replicate in practice than simple product features – with complaints volume, cost disclosure practices, and data governance quality as indicators to monitor.

In an environment where response speed is converging across providers, the way a lender manages borrower data becomes one of the last truly differentiating aspects.

Set against a market that has just recovered

Consumer lending in Vietnam expanded by around 26% in 2025, according to data from FiinGroup, with banks outpacing consumer finance companies by nearly double (VnEconomy, 2026). Material improvement, Asset quality: Consumer finance companies’ non-performing loan ratios fell from 11.2% peak in 2023 to 7.2% in 2025 and credit costs dropped from 14.5% to 11.5%.

Provider Scale as advertised by providers

As of May 2026, FE Credit had served nearly 17 million customers with more than 28 million contracts and over 20,000 points to refer services (up from its reported figures of more than 14 million customers with over 13,000 points in 2025) (Dân trí, 2025). For example, Home Credit Vietnam reported over 17 million customers and around 16,000 points of sale.

Both figures are as published by the company, which is why we have marked those [P] in the report rather than presenting them as independently verified.

Put differently, the sector entered 2026 with restored lending capacity and expanded reach – it was facing a new set of constraints deep down in the data layer.

What to watch through 2027

Three questions will determine how this plays out.

• If underwriting guidelines shrink (or not) as lenders finalize their compliance adjustments.
• If access for thin-file borrowers improves, declines or is unchanged.
• If the sandbox framework pursuant to Decree 94/2025/ND-CP (allowing for a testing system of up to two years) has created a permanent regime in alignment with the data protection regime instead of alongside it.

Vaynhanh Research does not make predictions about the outcome. Alternatively, recovering volume data regimes that are material in their changes cannot be the speed of disbursement for a competitive advantage. But most likely, it will be the ability to evaluate which borrower is the right fit for which product, and to show how that evaluation was done with lawful sunshine.

About the report

Vaynhanh Fast Loan Market Report 2026 confirmed sources on 31 July 2026. It uses categories of evidence: external fact, inferential calculation, estimated by the model, published by a provider-administrative dataset entry point, hypothesis-and-forecast; and ranks sources across five levels, with original official sources at their apex.

The full report is available at Vaynhanh.ai.