AI-integrated phone supports users to detect Deepfake calls - Laodong.vn

In late 2025, two separate incidents in Vietnam highlighted the growing danger of deepfake technology used to deceive individuals into transferring large sums of money. These cases involved sophisticated impersonations that successfully mimicked the appearance and voices of trusted figures, leading victims to believe they were communicating with legitimate contacts.

What happened

In late November 2025, Nguyen Thi Hanh, an accountant in Hanoi, received a video call from an account displaying the name and profile picture of her company leader. The visual and auditory quality of the call was so convincing that she did not suspect fraud. After completing a money transfer, she contacted the director via his personal phone number, only to discover she had been scammed out of more than 300 million VND.

In a separate incident in Bac Ninh, Ta Van Quang received a call from an unfamiliar number. The caller used a voice identical to that of his son, who was studying abroad in China. The person claimed to have been in an accident and requested an urgent transfer of 150 million VND for hospital fees. Due to the familiar tone and manner of speaking, Mr. Quang transferred the funds before realizing the interaction was a scam.

These incidents utilize AI to synchronize the voices and images of individuals, often harvesting data from social networks to increase the authenticity of the fraud. Beyond impersonating relatives or colleagues, scammers are also known to pose as police officers or functional agencies to coerce victims into transferring money under the guise of legal trouble. To combat these threats, some mobile manufacturers are now integrating on-device AI detection features that analyze video calls locally to identify synthetic content without uploading personal data to external servers.