What it is
AI-enabled fraud uses generated voices, faces, video and text to impersonate real people — a CEO, a colleague, a relative — or to create fake identities and documents that pass basic checks.
How it works
- Criminals collect public audio and video of the person they want to copy.
- AI tools clone the voice or face, sometimes in real time.
- The victim gets a call, voice note or video meeting that seems genuine.
- An urgent, confidential request follows: a payment, a password, a data transfer.
Red flags
- Urgency plus secrecy, even from someone you know.
- A request to bypass normal approval steps.
- Slight delays, odd lighting or audio glitches on a video call.
- A family member in trouble who can’t be reached any other way.
A real-world example
In 2024, an employee at engineering firm Arup in Hong Kong transferred around US$25 million after a video call in which the CFO and colleagues were all deepfakes.
How to protect yourself
- Verify unusual requests through a separate channel you already know.
- Agree a family “safe word” for emergencies.
- Require two-person approval for new payees and large payments at work.
- Limit how much voice and video of executives is publicly available where possible.
If it happened to you
- Call your bank now using the number on your card. Speed is the biggest factor in getting money back.
- Stop all contact with the scammer and don’t send more money — including “fees” to recover losses.
- Keep evidence: screenshots, numbers, emails, transaction references.
- Report it to the police and your national fraud reporting service.
Step-by-step help by country: I’ve been scammed — what now?
The rules
Laws are catching up: the EU AI Act sets transparency duties for deepfakes, and payment regulators increasingly expect firms to detect AI-generated identity documents during onboarding.
Source: www.cnn.com
