Just the Facts

Five Facts on Deepfake Videos

By Emma Petasis
June 25, 2019 | Blog

Deepfake videos use an artificial intelligence system to replicate someone’s face to make a realistic video of the individual saying or doing something that never happened. And these videos are increasingly being used to target political candidates. Here are five facts on them.

Deepfake videos are difficult to distinguish from the real thing.

Deepfake videos have been produced of multiple 2020 candidates, including one of President Trump’s face on a parody video of Alec Baldwin and a video of Sen. Elizabeth Warren’s face over actress Kate McKinnon’s –  both taken from SNL skits and that look like the candidates themselves.

The deepfake movement began in 2017.

The name was coined after a Reddit user posted a series of fake celebrity pornographic videos under the username “deepfakes.” The user superimposed the faces of celebrities onto the faces of others to create an illusion through seamless movement and coloration. 



The Pentagon is developing technology to detect deepfake videos.

The Defense Advanced Research Projects Agency (DARPA) under the Department of Defense is working with various academic institutions to invent technology to debunk deepfake videos. Hours of footage of all 2020 candidates have been downloaded to create a library of facial expressions and head movements of each individual, which can be used to test and validate videos before they are shared.[1]Researchers hope to have this software complete by the end of the calendar year.

Congress is questioning if social media companies should be held liable for fake content on their sites.

The House Intelligence Committee held a hearing on June 13 to analyze the potential risks of this technology. Currently, social media companies are not required to monitor or remove fake content from their platforms under Section 230 of the Communications Decency Act. [2]

Each social media platform has its own policies regarding the spread of false information. 

An edited video of Nancy Pelosi went viral across social media platforms in May 2019 which slowed down footage and slurred her words. Twitter did not remove the video of Pelosi due to the absence of any policies that require them to take down fake content. Facebook, which did not have a policy about removing fake information either, announced they would not delete the false video of Pelosi but circulated articles about how the video was fake. YouTube is one platform that does have policies requiring them to remove fake videos, and the videos of Pelosi were removed. [3]


[1]https://www.cnn.com/2019/06/13/tech/deepfake-congress-hearing/index.html

[2]https://www.theverge.com/2019/6/13/18677847/deep-fakes-regulation-facebook-adam-schiff-congress-artificial-intelligence

[3]https://www.theverge.com/2019/5/24/18637771/nancy-pelosi-congress-deepfake-video-facebook-twitter-youtube

Join us

Stay up to date.

  • This field is for validation purposes and should be left unchanged.

Share: