(This video explaining how deepfakes work and how they are made has been reposted from The Quint’s archive in light of the Bharatiya Janata Party reportedly using the technology to create videos of the party’s Delhi President Manoj Tiwari in Hindi, English, and Haryanvi to reach out to voters from different linguistic groups ahead of the Assembly elections.)
Deepfakes. Very real looking fake videos of people doing or saying something they never said or did.
The word deepfake comes from ‘Deeplearning’ and ‘Fake’.
Deepfake technology is becoming more and more common with each passing day and given the spread of Artificial Intelligence, its getting easier and more accessible at the same time.
The technology is not only becoming more sophisticated, it’s getting more easily accessible as well. Creating deepfakes now is as easy as processing thousands of photos of a subject through a software.
Stanford University researchers are developing a software where typing out a transcript will change what the subject is saying. A Cornell University study, on the other hand says that Samsung researchers are developing a deepfake tool that can create deepfakes using just one image.
Now, deepfakes use Neural Learning – a type of machine-learning technique that learns from the data it receives. The more photos or videos you feed into the system, the better your outcome.
To try and curb deepfakes, Reddit has already put out a ban. Websites like Twitter and PornHub are developing specific algorithms to filter these out. SRI International, a research company based out of California has also been contracted by the US’ Defense Advanced Research Projects Agency (DARPA) to create a tool that points out morphed videos.
(At The Quint, we question everything. Play an active role in shaping our journalism by becoming a member today.)