AI-driven deepfakes and voice cloning used in fraud scams in Vietnam

What happened

Criminal groups in Vietnam use AI-generated deepfake videos, images and voice cloning to impersonate family members, business leaders or trusted figures and request money transfers.

Criminal organizations have researched and provide tools to bypass biometric authentication in online payments, creating challenges for banks and payment intermediaries.

The nTrust fraud prevention system detected nearly 63,000 new types of malware on mobile phones in Vietnam in 2025, with fraudsters distributing counterfeit applications to steal data or take control of devices.

Reported impact

Fraudsters use purchased personal data classified by profession, residence, income level and lifestyle habits, combined with information from banks, shopping, and utilities, to build scenarios matching victims' circumstances and increase deception success.

Index analysis

Personal data faces multiple risks when AI is deployed, including users inadvertently providing data in prompts, systems collecting excessive information, and data persisting in knowledge bases, memory logs and caches.

Risks arise when AI systems have overly broad access permissions, connect to external systems or use third-party services; data protection measures applied at single points leave large gaps.

Classification

Organizations

Cục An ninh mạng và Phòng, chống tội phạm sử dụng công nghệ cao (A05), Bộ Công an (regulator), FPT, Hiệp hội An ninh mạng quốc gia

Category

Fraud, scams and manipulation

Industry

Digital security

Country

Vietnam

Harm status

Actual harm

Intent

Intentional

Timing

Unknown

Harm types

Economic or property

Harmed

Consumers, Business, General public

Failure modes

Deepfake video, Voice clone, Deepfake image, Manipulation or dark pattern

AI systems

deepfake generation systems, voice cloning technology, tools to bypass biometric authentication, malware for mobile devices

Sources and evidence