Deepfakes

What it is

A deepfake is synthetic media. It may swap a face onto another person’s body, generate a new video of a real person from still photos, or invent a person who does not exist. Quality varies: some fakes are easy to spot; others are good enough to fool a hurried viewer on a phone screen.

This is general knowledge, not a scare statistic. You do not need a special tool to start taking care — you need a habit of pausing before you trust what you see.

How people get hurt

Criminals use deepfake video or audio to impersonate a boss, a vendor, or a family member and push a rushed payment or a secret. Others use non-consensual intimate imagery to extort or humiliate. Political and brand deepfakes can spread faster than a correction.

• A fake “CEO on a video call” asking finance to move money now.
• A doctored clip of a public figure used to stir panic or smear a business.
• A fake ID-style portrait used to open an account or pass a weak KYC check.

How to slow it down

Treat unexpected video as untrusted until you confirm it through a channel you already know — a phone number already in your contacts, a meeting you scheduled, a colleague sitting next to you. Look for clues (odd lighting, mismatched ear-rings or glasses, lip-sync slips, a stiff or waxy face), but do not rely on clues alone. Better generators keep closing those gaps.

• Call back on a number you already have, not the number in the chat or the email.
• Agree a code word with family and with anyone who can approve payments (see our guides).
• Prefer a second factor: a written request in your normal finance system, dual approval, a live in-person check.
• If a clip is circulating about your organisation, say so clearly and point people to a known official channel.