AI Voice Cloning Scams Surge Across Thailand

Imagine your phone rings one afternoon and the voice on the other end sounds exactly like your younger brother — the same tone, speech rhythm, and urgent manner you've known all your life. He says he's in a financial jam and needs help right away. Almost no one who hears a voice like that stops to ask who's really calling, because our instincts tell us to trust the voices of people close to us. That instinct is the gap the new generation of scammers exploits. Today AI isn't just a convenience; it's a tool that lets criminals clone the faces and voices of people we love. Even as authorities continuously crack down on call-center gangs, the frauds keep evolving. Scammers now use AI and deepfake technology to fake voices, images, and video, making their stories far more credible than old tricks. Data from the Ministry of Digital Economy and Society's AOC 1441 center shows many Thais have already received AI-generated scam messages via SMS and social media, meaning this danger is no longer distant — it's already on our screens and in our incoming calls.

The first pattern to know is voice cloning. Scammers take voice samples that people themselves post in clips on social media, feed them to AI to learn and mimic the voices of relatives, close friends, or celebrities, then call victims to ask for loans, OTP codes, or sensitive data. A real case cited by AOC 1441 involved a victim who got a call from someone claiming to be their younger brother. The voice was nearly identical to the real sibling and asked for urgent money. The victim transferred 79,000 baht, only discovering the fraud after contacting the real brother later and immediately reporting it to AOC 1441 to freeze the account. What's striking is that this victim wasn't careless or distracted; they simply believed the voice they heard — something every human does.

The second pattern is deepfake images and video. Scammers often pose as police, government officials, or agencies to threaten victims into transferring money or to create panic so victims can't think clearly. The highest-loss case in this dataset exceeded 11 million baht. It started with a scammer impersonating a well-known telecom company, claiming the victim's name was used to register a SIM linked to an online gambling site. The call was then transferred to someone posing as an Anti-Money Laundering Office (AMLO) official, who sent documents resembling a court order via Line and arranged a video call with a fake senior police officer. The victim was told to verify financial trails to prove innocence and was threatened with legal charges if they didn't cooperate. Driven by fear, the victim transferred a total of 11,497,998 baht before realizing they could no longer reach anyone. Looking deeper, the terror isn't just the technology — it's the carefully staged, multi-step narrative: the SIM issue, the gambling site, the enforcement agency, the official-looking documents, and finally the uniformed figure on a video call. Each step pushed the victim deeper into panic, and in that state fact-checking disappears. The more realistic the AI-made audio and video, the more it seals off the doubts a victim might have had.

The AOC 1441 center has compiled seven warning signs to help people recognize an attack in time:

A call comes in but stays silent, or tries to get you to...