--- title: "What is Deepfake? — Social Media Glossary | Postpone" canonical_url: "https://www.postpone.app/social-media-glossary/deepfake" last_updated: "2026-10-08T07:16:26.084Z" meta: description: "A deepfake is AI-generated synthetic media that convincingly replaces someone's likeness or voice in a video or audio recording." "og:description": "A deepfake is AI-generated synthetic media that convincingly replaces someone's likeness or voice in a video or audio recording." "og:title": "What is Deepfake? — Social Media Glossary" "twitter:description": "A deepfake is AI-generated synthetic media that convincingly replaces someone's likeness or voice in a video or audio recording." "twitter:title": "What is Deepfake? — Social Media Glossary" --- **Social Media Literacy** # Deepfake A deepfake is AI-generated synthetic media that convincingly replaces someone's likeness or voice in a video or audio recording. [**Reddit **post scheduler](https://www.postpone.app/platforms/reddit-post-scheduler) [** X **post scheduler](https://www.postpone.app/platforms/twitter-post-scheduler) [** Instagram **scheduler](https://www.postpone.app/platforms/instagram-post-scheduler) [** Facebook **scheduler](https://www.postpone.app/platforms/facebook-post-scheduler) [** LinkedIn **scheduler](https://www.postpone.app/platforms/linkedin-scheduler) [** TikTok **post scheduler](https://www.postpone.app/platforms/tiktok-post-scheduler) [** Threads **scheduler](https://www.postpone.app/platforms/threads-scheduler) [** Pinterest **scheduler](https://www.postpone.app/platforms/pinterest-scheduler) [** YouTube **Shorts scheduler](https://www.postpone.app/platforms/youtube-shorts-scheduler) [** Bluesky **post scheduler](https://www.postpone.app/platforms/bluesky-post-scheduler) [** Tumblr **scheduler](https://www.postpone.app/platforms/tumblr-post-scheduler) [** Mastodon **post scheduler](https://www.postpone.app/platforms/mastodon-post-scheduler) Deepfakes use artificial intelligence to create video, audio, or images that look and sound real but are entirely fabricated. The technology works by training neural networks on existing footage of a person, then generating new content that maps their face, voice, and mannerisms onto different source material. ## How They're Made Early deepfakes required significant computing power and technical skill. That's no longer true. Free and low-cost tools can produce convincing face swaps in minutes. Voice cloning services need as little as 30 seconds of sample audio to generate a passable imitation. The barrier to entry drops every year. The quality range is wide. Cheap deepfakes have telltale glitches — unnatural blinking, warped edges around the face, audio that doesn't match lip movements. High-end deepfakes produced with better tools and more training data are nearly indistinguishable from real footage to the human eye. ## The Damage The most common harmful use of deepfakes is nonconsensual pornography. Early research estimated the vast majority of deepfake content online fell into this category, overwhelmingly targeting women. While the share of non-pornographic deepfakes has grown with mainstream adoption, it remains a major form of image-based abuse that's become trivially easy to produce. Beyond that, deepfakes are used for: - **Financial fraud** — impersonating executives in video calls to authorize wire transfers - **Political manipulation** — fabricated speeches or statements designed to go viral before elections - **Scams** — cloning a family member's voice to call and request emergency money transfers - **Reputation attacks** — placing someone in fabricated compromising situations In 2024, a finance worker in Hong Kong reportedly transferred $25 million after a video call with what appeared to be the company's CFO — every other person on the call was allegedly a deepfake. ## Detection and Defense Platforms are building automated detection tools, but it's an arms race. Detection models trained on today's deepfakes won't catch tomorrow's improvements. Some approaches focus on digital provenance — cryptographic signatures that verify where and when media was originally captured. For individuals, the best defense is skepticism toward viral video and audio, especially when it shows a public figure saying something shocking or out of character. Check if the original source is verified. Look for reporting from multiple outlets. If a video only exists as a screen recording reposted on one platform, treat it with extreme caution. The deeper issue isn't just fake content — it's the erosion of trust in real content. When anyone can claim a genuine video is a deepfake, the concept of video evidence loses its power. That uncertainty benefits bad actors regardless of whether they actually use the technology. **Example** A deepfake video of a CEO announcing fake earnings circulates on Twitter/X and gets 2 million views before it's debunked. The company's stock drops 4% in the hour before the video is confirmed as fabricated. ## **Why It Matters** Deepfakes make it possible to put words in anyone's mouth with alarming realism. They've been used for financial fraud, political manipulation, and nonconsensual pornography. As the technology becomes cheaper and more accessible, the ability to trust video and audio evidence erodes for everyone. ## **Related Terms** [Astroturfing](https://www.postpone.app/social-media-glossary/astroturfing) [Going Viral](https://www.postpone.app/social-media-glossary/going-viral) [Bot / Social Bot](https://www.postpone.app/social-media-glossary/social-bot) [Trending](https://www.postpone.app/social-media-glossary/trending) [**← Back to Social Media Glossary **](https://www.postpone.app/social-media-glossary)