DNA Special: What is a 'Deepfake' and why this technology may be dangerous?

In today's show, we will tell you about a threat, which can pose a bigger challenge to the world than terrorism, and this threat is from the Deepfake video - a technique that can be used to discredit anyone by making a fake video in a few hours.

To understand this, first, you will have to watch some fake videos, which are quite viral on social media. Some people are enjoying these fake videos, but there are many people who are worried. And to understand what these people are worried about, you have to watch these videos first.

This is Deepfake technology, which manipulates real photos or videos with the help of Artificial Intelligence, turning them into fake photos and videos for a specific purpose. And all this is done so cleanly that it becomes difficult to distinguish between real and fake pictures and videos. This is why it is called the Deepfake technique.

It can happen to big celebrities, it can happen to the prime minister and ministers of any country and it can happen to big industrialists and common people too. Think how dangerous this technique is.

Such Deepfake videos are often spread with the help of social media and are at the forefront of Facebook but Facebook founder Mark Zuckerberg himself has fallen prey to Deepfake.

Some time ago, one of his fake videos went viral and even then it was difficult for people to understand whether this video is real or fake. Not only this, this has also happened with former US President Barack Obama.

Now you must have understood why we are saying that this technology can become a bigger challenge for the world than terrorism. To understand this, now you have to look at some figures.

Every day, 180 million photos are uploaded on social media all over the world, i.e. the number of photos uploaded on the internet in a week is equal to the population of the world at this time.

Crores of these pictures are in the form of selfies and the number of pictures of women is much more than that of men. If you are a woman, today, you may have posted a picture of yourself on social media, but do you have any idea what cybercriminals can do with your pictures?

With the help of Deepfake Technology, these pictures can be easily converted into pornographic images. And you can understand this through a study done in 2020.

This study was done by a company that found fake content on the Internet, according to which, photos of one lakh women put on social media have been converted into pornographic images with the help of this technique.

How Deepfake technology works

With the help of Artificial Intelligence (AI), some facial changes are made to create a fake video. And many times, the voice is copied from the original video and put into a fake video. And this kind of video feels like real video. You can also call such a video morphed video, doctored video, or synthetic media.

How you can avoid Deepfake videos

Videos are easier to identify than Deepfake photos. And you can do this with the help of two things. For example, when a Deepfake video of a person is made, there is not much difference between that person and the background. But if you recognize that the focus is only on the face in the video and the background is deliberately hidden then you can identify the fake video.

The second way is to develop such cameras in future, which will take pictures and convert them into a kind of digital signature and make them encrypted. Due to this, criminals will not be able to use them easily.

An easy way to protect yourself from Deepfake is to share a minimum number of photos on social media and avoid putting close-up pictures of your face as much as possible.

The fourth way is to develop Artificial Intelligence which recognizes Deepfake quickly and ensures its reach to the common people so that people can avoid falling prey to fake news and fake videos.

Most of the photographs which are easily tampered with are Selfies.

In Selfies, the face is clearly visible and generally, its resolution is also good, so it becomes easy to associate a face taken as a selfie with a pornographic photo. It is very dangerous to put a close-up of your face on social media.

Apart from this, if you are fond of taking selfies from different angles, then you should also abandon your habit because these different angles of your face make the task of Artificial Intelligence preparing Deepfake very easy. With this, videos can be produced which are very difficult to believe to be fake, because it contains almost all the angles of your face.

According to a report in the year 2019, 14,000 such videos were identified on social media, which were produced with the help of Deepfake software. And the important thing is that 96% of these videos were related to pornography.

Why is the term Deep Fake used for such videos

In the year 2017, there was a user named Deep Fake on a website named Reddit and this user then started using facial change technique for pornography. It was from here that the word Deepfake came to the world and such videos came to be called Deepfake Videos.

However, there is another side to this technology and if it is used properly, with the help of this, many changes can be brought.

For the past few days, a software named MyHeritage is in news. With the help of this software, you can convert any picture to a 10-second video. With this software, you can add life to old photographs too.

Pictures of Shaheed Bhagat Singh, Swami Vivekananda, Lok Manya Tilak, famous Hindi writer Munshi Premchand, Lal Bahadur Shastri and famous French ruler Napoleon Bonaparte have been converted into videos with the help of this software. And these videos are enough to tell that this technique can be useful for us if Artificial Intelligence is used in the right direction.