Feb 21, 2020

Hany Farid Predicts a Troubling Future as Deepfake Technology Evolves

From NBC Bay Area

Deepfakes: Election Interference Just The Start of Threats Posed By Evolving Video Technology

By Candice Nguyen, Josh Slowiczek, Sean Myers, Michael Horn and Jeremy Carroll

We all recognize them: Lincoln standing for a portrait, Lenin addressing Russian soldiers, a woman crouched beside a student's body on the ground at Kent State University in 1970. But did you know those canonized points-in-time were altered?

Photo manipulation has been around for almost as long as the technology itself, a practice reserved for a few particularly-skilled individuals around the world. But that is all changing with the use of machine learning in video production, technology better known as deepfakes. Now, almost anyone can download the computer code and programs required to create digitally-altered videos of whoever, wherever or whatever they want. That is, if you know where to look.

“I’ve been thinking about these problems in my lab for about 20 years now,” said Hany Farid an expert on digital authentication teaching at UC Berkeley. “We have more and more examples of manipulated media being used to disrupt democracies, sow civil unrest, revenge porn, spectacular amounts of disinformation that lead to health crisis, and so on and so forth.”


Professor Hany Farid has a joint appointment at UC Berkeley in the School of Information and EECS.

Last updated:

February 27, 2020