What is DeepFake technology?
Deepfake is a product of Artificial Intelligence technology; It has frequently been used to create and propagate fake video content and generate what is known as ‘fake news’. It was first noticed in 2017 when a Reddit user altered pornographic video content by placing celebrities faces onto them.
It is a combination of two clashing AI technologies: the generator and the discriminator. The generator is the creator of the fake content while the discriminator is the filter or the identifier that decides if the content is fake or authentic. These technologies are adaptive and self-learning since they learn from their experiences. The discriminator provides insight into the characteristics that make the content seem unauthentic each time while the generator adapts and uses these learnings to take a different approach. The combination of these two systems makes up the generative adversarial network (GAN).
Implications of Deepfake Technology
Deepfakes can be used to create positive and negative effects. It has been used in the past to help ALS patients or people with speech impairments to communicate effectively. It can also be used to create efficiencies in the film, medical and non-profit industries through major cost savings. However, there are several negative uses as well. Video and audio have long served as proof of authenticity.
With the change in dynamics due to Deepfakes, this authenticity has been challenged and people have no awareness of the possibility of tampering or its ramifications. World leaders, corporate heads and celebrities can now easily be framed saying things that were never said leading to catastrophic consequences. With that being said, not much has been done from a legal standpoint to monitor creation or distribution of this AI monstrosity.
Damage to Enterprises
One of the biggest damage to enterprises is misrepresentation in the form of fake statements made by company. With Deepfake technology, this is a real possibility. Deepfake technology is now being used to launch almost unidentifiable phishing attacks where it is difficult to differentiate between the real and fake entity.
Fake content can arise from within as well for instance, in the form of fake reviews to meet important Key Performance Indicators. These practices must be governed by cyber security policies dictating use of company IT assets.
Measures against Deepfake
A block-chain can serve as an effective way to counter tampering of videos. It is a digital ledger that documents any modifications made to an original video so that the creator can track changes.
Even though using a discriminator to identify a fake video is a possibility, since GAN is a self-learning network, it will begin to produce even better fakes with this learning.
Hashing is another technique which is essentially a digital watermarking that provides each video with a combination of numbers that is unique to the video and is lost if the video is modified.
In the US, social media platforms have majorly user driven content and are not held responsible for the content shared on their platforms. However, Congress is considering changing this and making such immunity dependent on “reasonable moderation practices”.
Companies need measures to monitor mentions of the company and its employees over the internet. They can work in tandem with content delivery networks to manage or at least monitor this information. The most effective way to prevent misrepresentation of an enterprise is to secure your identity
Securing your Identity
It is imperative that organizations realize the need to implement strong identity protection measures. These measures need to not only capture employee bio metrics but also determine how best to safeguard the information so that it can’t be used to create potential fakes.
The world of technology is evolving and we must evolve with it. There will be a time when it will be impossible to distinguish real from fake content and at that stage, it is important that our credibility as an individual, enterprise and nation is intact. The reason behind this is that people will only judge on the source and context of the content and not the quality. This means that the best practices we employ today to secure our identity will go a long way in building and maintaining our credibility for now and the future.