How genuine is the internet? That question may seem redundant, given the recent turn of global events. We come across a minimum of ten photos or video that surprise or amuse us, but eventually turn out. All generated using cutting-edge AI. The Deepfake technology that is based on Generative Adversarial Networks (GANs) was invented in 2014 by a PhD student, Ian Goodfellow who now works for Apple. When two GANs communicate to the other about an image being ‘real’, the algorithms are enabled to generate or create those images. An advanced GAN is capable of creating a video clip of a person just with a single image.
Most of the online content produced through the Deepfake application is pornographic. The availability of the app in the hands of unskilled people makes it easier for them to create any number of fake videos and photographs. Women being the target, most times.
The challenge of Deepfake technology is not just about videos, but it allows users to create fictional photos from nothing. The technique has triggered an increase in the number of fake profiles on LinkedIn and Twitter for spying, spamming and scamming. The technology also allows the faking audio which results in voice skins or voice clones of public figures being circulated on social media platforms. Several scams have been carried out using such audio clips during phone calls.
Disadvantages of Deepfake
There are websites like Deepnudes, where faces of women including celebrities and public figures could be overlapped on pornographic video content. The advancement in Artificial Intelligence has resulted in possibly anyone and everyone making such fake content, feeding into disinformation.
Altering video footage wasn’t impossible for decades, but wasn’t this east either. The widespread accessibility of the technology has resulted in a state where fake videos are created by anyone including state-sponsored actors, political parties or lone individuals. The technology may lead to a serious risk to national security if unless people are able to discern the fake from genuine. Easier said than done eh?
There could more Deepfakes in the nearest future that are meant to intimidate, demean or harass an individual, group or nation. They may not cause any major international incident or trouble as most nations have their reliable security imaging systems. But this will severely dent trust in society.
Is Deepfake a boon at times?
Deepfake technology is a huge threat to society in several ways, but it can sometimes be entertaining and helpful. The voice of people who lose them due to diseases can be restored with the help of Deepfaking. Videos can be used to recreate museums, art galleries or even dead actors in movies.
Deepfaking can help to increase the popularity of a web page of profile popular on search engines as more people start searching for topics that attract their attention.
The issues over fake images are going to increase, as Deepfake technology continues to grow. There are several ongoing research projects being conducted to help detect fake images and videos and control further circulation.
And that bring us back to square one. Deepfake – a boon when put to sensible use, or a bane in the hands of temptation and mischief?