Deepfake phone and video scams
AI-generated audio and video is being used to impersonate executives, family members and celebrities in real-time calls. Deepfake vishing attacks rose 442% between the first and second halves of 2024. These scams are now reaching UK businesses at scale.
What are deepfake scams?
Deepfakes are AI-generated media (audio, video or images) that realistically depict a real person saying or doing something they never said or did. Fraudsters use deepfake technology in phone calls, video calls and social media to impersonate trusted people and extract money or sensitive information.
The technology requires only a small sample of someone's voice or face — often scraped from publicly accessible social media — and has become significantly cheaper and more accessible since 2023. What previously required expensive studio equipment can now be run on a consumer laptop.
Real cases
Arup: £25 million, 2024
British engineering firm Arup lost approximately £25 million after a finance employee was deceived by a deepfake video call in which AI-generated versions of company executives, including the CFO, who instructed transfers to fraudulent accounts. The employee believed the call was genuine throughout.
WPP CEO impersonation, 2024
The CEO of advertising group WPP, Mark Read, was impersonated using a deepfake voice in a WhatsApp video call that also used a publicly available photo as the caller's avatar. The fraudster used the fake identity to solicit money and personal information from a senior executive.
Hong Kong finance worker: £20 million, 2024
A finance employee at a multinational firm was tricked into transferring £20 million after joining a video call where all other participants — including the company's CFO — were AI-generated deepfakes. The call appeared authentic throughout.
Known attack patterns
CEO / CFO fraud via video call
A video call is staged where an AI-generated version of a company executive instructs a finance employee to make an urgent wire transfer. The employee, seeing and hearing their executive, authorises it.
Investment scams using celebrities
Deepfake videos of well-known investors or celebrities appear to endorse fraudulent investment schemes. Victims are called and the video is used to build credibility for the fraudster.
Family emergency scams
An AI-generated voice or video of a family member calls claiming to be in trouble: arrested, hospitalised or stranded abroad. They request an immediate bank transfer.
Romance scams via video call
AI-generated video calls maintain a fake romantic relationship over weeks or months. The victim believes they are in a real relationship and eventually sends money when the fake person presents a crisis.
How to detect deepfakes
- →Look for unnatural blinking, facial movements or lip sync that does not quite match the words.
- →Audio quality may be slightly inconsistent or robotic at certain moments, particularly on less-common words.
- →The background may appear blurred or distorted around the face edges.
- →Ask an unexpected question or reference something very recent, as AI systems struggle with spontaneous, specific responses.
- →Establish a video call challenge with colleagues: ask them to turn their camera off and on again, or perform an unusual gesture.
- →If the call is genuinely unexpected, hang up and call back on a number you already have saved.
Protection: for individuals and businesses
- ✓Never authorise financial transfers based solely on a phone or video call. Always verify through a separate, established channel (call back on a known number).
- ✓Establish code words with family and colleagues for use in suspicious calls.
- ✓Businesses should implement dual authorisation for all significant financial transactions. One call or video message should not be sufficient.
- ✓Be cautious about how much video and audio of yourself is publicly accessible online.
- ✓Treat unexpected urgency as a red flag: genuine executives and family members will not mind you taking two minutes to verify.