Last checked: 7 November 2026
In short: A deepfake is a fake video, image or voice recording made with AI to show a real person saying or doing something they never did. Modern tools can produce a convincing fake in minutes.
Table of Contents
How it works
- Gather samples. The creator collects photos, video or audio of the person, often from interviews and social media.
- Copy the person. AI software learns the person’s face and voice, or a ready-made tool copies them from a short clip.
- Produce the fake. The tool generates new footage or speech, sometimes live during a call.
Where you’ll meet it
Voice cloning in tools such as ElevenLabs, and video generators such as Google’s Veo, can create realistic people. Fakes then spread through WhatsApp, Microsoft Teams, Zoom and social media.
What it means for your business
- Seeing is no longer proof. A familiar face on a video call does not confirm who is speaking.
- Change payment controls. Require a call-back on a known number, or a second approver, for any urgent transfer or change of bank details.
- Train your staff. Show finance and HR teams real cases so they know to pause and check when a request feels rushed.
Watch out for
- Executives are easy to copy. Public talks and podcasts give fraudsters plenty of voice and video to work with.
- Detection tools miss fakes. Software that claims to spot deepfakes is often wrong, so rely on process checks.
Example
Arup. In January 2024 an employee in the engineering firm’s Hong Kong office joined a video call with what looked like the chief financial officer and other colleagues. All of them were deepfakes. The employee then made 15 transfers totalling HK$200 million (about US$25 million) to fraudsters. In May 2024 Arup confirmed that fake voices and images were used.
Deepfake vs misinformation
| Deepfake | Misinformation | |
|---|---|---|
| What it is | AI-made fake media of a real person | Any false or misleading information |
| Form | Video, image or audio | Text, images, video or speech |
| How they relate | One tool for spreading falsehoods or fraud | The wider problem, with or without AI |
Frequently asked questions
Are deepfakes illegal in the UK?
Some are. Using one for fraud is a crime, and sharing sexual deepfakes of real people without consent is an offence. Clearly labelled satire and consented uses are generally legal.
How can I tell if a video call is fake?
Often you cannot by looking. Ask a question only the real person would know, or end the call and ring them back on a number you already hold.
Can my business use deepfake technology legitimately?
Yes, with consent. Firms use AI avatars and cloned voices for training videos and translations, as long as the person agrees and viewers are told it is synthetic.