A video of Rashmika Mandanna, one of India’s most famous actresses, has gone viral on numerous social media platforms recently. The actress in that video was seen wearing questionable attire as she entered an elevator. The popularity of the video led to widespread belief among social media users and the public that the individual in the video was precisely Rashmika Mandanna. Her dressing sense became a source of trolls for many. After the video went viral, it was revealed that it was not real and was created with deepfake AI technology. The actual woman in that video was Zara Patel, a British-Indian influencer. The video was altered by using deepfake technology to change Zara Patel’s face to Rashmika Mandanna’s. Deepfake technology has become increasingly prevalent over the past few years. A reactive and controversial issue was highlighted after this viral video. Prior to this viral video featuring Rashmika Mandanna, recall the Zomato ads on YouTube where Bollywood actor Hrithik Roshan spoke about a popular restaurant in your area. The specialty of that advertisement was that the name of the restaurant and the location changed depending on the viewers’ current location at that time. The same case happened in the advertisement for Shah Rukh Khan on Cadbury. He promoted different Cadbury stores in different locations, depending on the location of the viewer who was watching the ad video.
Although the deepfake tool or AI technology we’re talking about isn’t new, it was brought to light after it was misused in Rashmika Mandanna’s viral video. The social-cultural system can be negatively affected by fake technology, as shown in this single video. Social media platforms have been asked by the Union IT minister of state, Shri Rajeev Chandrashekhar, to take prompt action against fake videos. Accordingly, he said that “under the IT rules notified in April 2023, it is a legal obligation for platforms to ensure no misinformation is posted by any user and ensure that when reported by any user or government, misinformation is removed within 36 hours.”
What Is Deepfake Technology
The name Deepfake comes from the combination of the words ‘deep’ and ‘fake’. Deep machine learning is a type of artificial intelligence tool or technology. With the help of this technology, you can easily manipulate or modify an existing image, video, or voice of someone or something. Facial recognition is the fundamental concept behind this technology. It’s common to discover that massage apps or Snapchat have multiple filter tools that enable you to quickly alter or modify your facial expressions. Deepfake is similar to this, but at an advanced level. This technology creates images or videos that are much more realistic, making it very difficult for an ordinary person to differentiate them from the original ones.
Types Of Deepfake
Deepfakes are mainly classified into three types. Which are as follows:
(1) Face Swap
This is a most popular deepfake technology where you can change the face of a person without any modifications of his/her body.
(2) Facial Attributes and Facial Expression Manipulation
Applying this allows you to alter a person’s skin color, age, gender, and even their facial expression.
(3) Face Synthesis
With this advanced technology, you can create a completely new person who does not exist in the real world. There is a website called ‘ThisPersonDoesNotExist.com’. On this website, every time you refresh the page, it will show you an image of a new person who does not exist in the real world. All images were created using DeepFake technology.
Beyond the generation of fake photos or videos, deepfake AI technology can also create fake audios of any person, and the best examples of this were considered in the advertisements of Zomato and Cadbury previously.
How Does Deepfake Technology Works ?
Two neural network systems known as ‘encoder’ and ‘decoder’ are necessary for deepfake technology. A neural network, or artificial neural network, is an artificial computing system inspired by a biological neural network. A neuron present in our body works like a messenger, which, with the help of electrical impulses and chemical signals, shares information between our brain and nervous system. Deepfake is also a neural network-based technology that creates suitable results on the basis of the received information. This tool can easily perform creative tasks such as facial rejuvenation and audio replacement. The first task of deepfaking is to break the image into small details, which is also called compressing. Deepfake uses another neural structure to reconstruct this compressed image again. Encoder-decoder models are used to manipulate attributes such as face swapping and facial expressions.
Another tool for deepfake is the Generative Adversarial Tool, or GAN. This tool is used to create non-existent images. To create a deep fake, much more data is fed into the AI model to train it. This process is called machine learning. With machine learning, a non-living biological body can learn like a human with the help of data and algorithms.
Creating a convincing deepfake video is a time-consuming task. But with the advancement of AI technology, this issue can be resolved.
Impact of Deepfake Technology on Human Society
Benefits of Deepfake Technology
Deepfake technology has the potential to be an efficient and economical advertising tool. Customers are drawn to that specific company by the highly tailored experience that customized deepfake advertisements offer. According to some specialists in digital marketing, businesses will eventually be able to buy a license to use an actor’s identity in advertisements. With this technology, the entertainment business of the future can greatly benefit. Deepfake can be used to dub original films in a variety of regional languages. It can also improve the experience of watching movies and lower the cost of employing local voice actors. additionally resolve the issue of actors’ unavailability. It may be useful when determining an actor’s age for a film.
Opposite Sides Of Deepfake Technology
It can blur the lines between truth and lies. can have the chance to believe lies and disregard the truth. Deepfake can increase the problem of the existence of trust in this fake world. The lack of digital literacy in many countries around the world, like India, makes deep fakes even more dangerous. Most people do not check if a video or image is true or false before sharing it on social media, which will decrease a person’s social privacy. The most dangerous thing is that fake videos can be a potential threat to international relations and democracy. It has a chance to increase financial fraud and cybercrime. As per the data, 96% of non-consensual pornographic contents on the internet are fake. Assume an Artificial Intelligence technology that can manipulate or alter an image, video, or voice. How dangerous will it be if used with bad intentions at a time when human society has lost its tolerance and patience? Now a days, a small oral controversy between two persons relating to ideology, religion, caste, and general belief can generate mob violence in a few hours.
Conclusion
Companies including Adobe, Google, and Microsoft are investing in the development of deepfake detection technology. Yet, Microsoft’s chief scientific engineer, Mr. Eric Horvitz, claims that although deepfake detection technology is advancing, advanced deepfake technology is expanding concurrently, and these two will continue to compete in the future. Therefore, identifying the authentic video, image, or voice in this case is more crucial than spotting the fake one. Microsoft and Adobe have developed a tool called content credentials to ascertain the legitimacy of a picture or video in order to address this issue. An authentication tick will appear along with an original video or image when it is shared on social media. The visitor can access the specifics of that image or video by clicking the authentication tick. On the other hand, the government ought to enact stringent laws to combat deepfake and should also develop awareness campaigns about its negative effects on the sociocultural environment across all print and electronic media channels.
********
To read many more interesting articles, please visit our website below 👇