In plain English
A deepfake is synthetic video, audio, or imagery created with deep learning to convincingly depict a real person saying or doing something they never did. As the technology has improved, casual inspection often fails to catch the better ones. The same techniques have legitimate creative uses, but the term is most associated with fraud and misinformation.
Why it matters
Deepfakes erode the assumption that audio and video are reliable evidence, enabling scams, harassment, and political manipulation. Verifying the source of media matters more than ever.
A worked example
A scammer clones an executive's voice from a few seconds of audio and calls the finance team to authorise an urgent payment that never should have happened.
Common confusion
The absence of obvious glitches no longer proves a clip is real. High-quality deepfakes can look and sound flawless, so context and provenance matter more than visual tells.