When picking up a video call that seems to be from a family member, will you suspect that it may be a scam using Artificial Intelligence (AI) technology? We used to say that “pictures don’t lie”. However, in the age of AI, pictures, sound clips, videos or even live video calls may not be true.
AI has witnessed unprecedented advancements in recent years. Scammers also make use of the technology to commit fraud, e.g. using Deepfake and voice cloning to replicate a person’s appearance and voice, creating deceptively real videos or voice. Voice cloning is a common AI technology that scammers employ to imitate someone’s voice based on just a short audio sample while Deepfake uses AI synthesis to fabricate extremely convincing facial expressions and speech.
Common tactics used by scammers include impersonating a victim’s family member, friend or boss and asking for money transfer or confidential information, or posing as celebrities or investment experts to lure victims into investing in virtual assets. Since both the images and voices are so realistic, it is very difficult to tell what is real, making it easy to fall into the traps.
We should always stay vigilant, especially when being offered an attractive investment opportunity or receiving any requests involving money. Think twice and verify the facts carefully, such as calling the related person directly to confirm the request. You may also take note of the Police’s tips on preventing Deepfake scams:
- Ask the caller to do specific gestures in front of camera and watch carefully for any anomalies;
- Speak in another language and see how the caller may react;
- Use a pre-agreed secret code word to confirm the caller’s identity;
- Verify any money transfer requests through multiple channels;
- Avoid answering video calls from unknown numbers;
- Evaluate fraud risks using “Scameter”;
- If in doubt, call the Hong Kong Police’s Anti-Scam Hotline at 18222 for assistance.
