Deepfake is synthetic media — usually video or audio — in which AI is used to convincingly replace or fabricate a person’s likeness or voice, making them appear to say or do things they never did.
What it means in plain English
The word combines “deep learning” and “fake.” Using AI, a deepfake can swap one person’s face onto another’s body, or clone a voice to generate speech that person never spoke. As the technology has improved, deepfakes have become increasingly realistic and easier to make, raising serious concerns about misinformation, fraud, and consent.
Not all synthetic media is malicious — the same techniques have legitimate uses in film and accessibility — but the term usually carries a warning.
A simple example
A deepfake video might show a public figure appearing to say something inflammatory they never actually said, created by training an AI on real footage of them and generating new, fabricated speech and expressions.
Why it matters
Deepfakes are one of the most concerning misuses of generative AI, with real implications for trust, security, and democracy. Understanding them — and that convincing fake media can be manufactured — is increasingly important media literacy in the AI age.
Related terms
- Generative AI — the technology behind deepfakes.
- GAN — historically used to create deepfakes.
- Watermarking — a technique to help identify synthetic media.
Frequently asked questions
What is a deepfake?
A deepfake is synthetic media — usually video, image, or audio — in which AI is used to convincingly replace or generate a person’s likeness or voice.
Why are deepfakes a concern?
They can be used to deceive, spread misinformation, harass, or commit fraud by making people appear to say or do things they never did, which raises serious consent, trust, and legal issues.
