Big Update Deepfakes: Why has social media been summoned by the government of India in 2023?

As digital technology is progressing, along with its benefits, it also faces its disadvantages. Numerous kinds of wrongdoings are being done by fraudsters by using technology. One of them is misuse of deepfakes. In this article we will explore what are deepfakes and why has social media been summoned by the government of India regarding it.

What are Deepfakes?

Deepfake means “deep learning” + “fake”. Deepfake works on a technology called ‘Human Image Synthesis’. Machine Learning and AI Technology are used in this process. It can be Photos, Videos or even Audio. It had been developed by the researchers at academic institutions in the starting of 1990s and thereafter it was adopted by amateurs of online communities and industries.

Just as we can photocopy anything, similarly this technology can make exact copies of people walking and speaking. Meaning, we can see any person or creature moving or speaking on the screen, which is fake or whose copy has been made.

With this, the face of one person can be superimposed on the face of another person and that too with such precision and detail that all the expressions of the source face can be seen on the fake face. Apps built on the foundation of this technology can cause immense harm.

Why has Social media been Summoned by the Government of India?

Deepdeck videos are being shared on social media by misusing AI. Keeping in mind the misuse of deepfake technology, the Indian Government has adopted a strict attitude against social media platforms.

The central government is in action mode on this issue and the government has sent notices to all the major social media companies.

Indian Government has sent a notice to all social media platforms to immediately remove deepfake photos or videos available on their respective platforms and take immediate action against them. If the social media platform itself remains ineffective then whatever reliefs have been provided by the government will be removed.

On this subject, a meeting is also going to be held by the government along with social media platforms in the next 3-4 days in which the pitfalls of deepfake will be considered and efforts will be made to find solutions to prevent it. The government is also trying to bring laws to stop the misuse of AI.

Let’s know something more about deepfakes –

How Does Deepfake Work?

As I explained above, deepfake technology can be used to create an exact replica of a person or object using “Machine learning” and “AI Technology”. In this process, original photos or videos are given to computers and on giving commands; it creates its fake copy as per our desire. This is almost like a child seeing something and making a drawing of it.

These fake photos or videos have hidden layers. These look absolutely genuine and their fakeness can be identified only through editing software.

Deepfakes are created with the help of an encoder network and the decoder network. The encoder network analyzes the source content (the original video) and then sends the data to the decoder network. Final output which comes out is exactly like the original but in reality it is fake.

You may also like:

• What is D2M Technology in 2023?

• Which Device May Protect Your Phone From Juice Jacking

Types of Deepfakes:

Deepfake may be categorized in the following manners –

• Face-swaping: In which the face in the video is automatically replaced with another person’s face.
• Lip-syncing: In which a source video is modified so that the mouth and lips of the person speaking correspond to an arbitrary audio recording.
• Puppet-Mastering: In which an actor mimics (head and eye movements, facial expressions) a target person seated in front of a camera.

What are the Examples of Deepfake?

• Recently a fake video of South Indian actress Rashmika Mandana was viral. In this video, a girl’s face has been replaced with the actress’s face. This video looked absolutely original.
• What’s more, recently Prime Minister Shri Narendra Modi said that he saw a fake video of himself in which he was playing Garba.

In addition to the above there are numerous examples of using deepfake technology all over the world especially in politics. Some examples are as follows –

• A deepfake of Barack Obama featuring Jordan Peele’s voice was produced in April 2018 as a public service message to raise awareness of deepfakes. Peele and Buzzfeed collaborated on the project.
• In two different videos released in year 2018, the faces of Angela Merkel were swapped with Donald Trump’s and the face of Argentine president Mauricio Macri with Adolf Hitler.
• A deepfake of Trump during his Oval Office speech was broadcast by Fox affiliate KCPQ in January 2019, making fun of his appearance and skin tone. The worker who was later identified as the creator of the video had been fired.
• Similar technology was utilized by the Delhi Bharatiya Janata Party during the year 2020 Delhi Legislative Assembly election campaign to disseminate a Haryanvi translation of an English-language campaign ad by its leader, Manoj Tiwari, in an attempt to win over Haryana voters. An actor delivered the voiceover, and the film was lip-synchronized to the new voiceover using artificial intelligence (AI) trained on Tiwari speech videos. It was a “positive” application of deepfake technology, according to a party staff member.
• A deepfake video purporting to show Vice President Kamala Harris slurring her words and talking incoherently about today, tomorrow, and yesterday went viral on social media in May 2023.
• In order to mislead people about Donald Trump, Ron DeSantis’s presidential campaign utilized a deepfake in June 2023 in the USA.

Is Deepfake a Cybercrime?

In India Threat actors are using deepfakes, which are quickly becoming an issue, to disseminate false material online. Nonetheless, threat actors can be discouraged from producing deepfake videos by using certain legal measures.

According to India’s IT Rules, 2021, intermediate platforms must remove any content that is found to be fraudulent or created using deepfake within 36 hours.

The Indian IT ministry has also sent letters to social media networks indicating that impersonating someone online is prohibited by Section 66D of the Information Technology Act of 2000, following the widespread success of Rashmika Mandanna’s deepfake videos.

The IT Rules, 2021, also mandate that social media companies remove photos that have been artificially altered upon being notified, and they forbid hosting any content that impersonates another person.

Why do Cybercriminals Use Deepfakes?

Deepfake technology is used by some people to make fun of others. Still, fraudsters also create fake photos or videos to defame some popular celebrities.  Moreover, this technology is also used to blackmail a person by creating fake obscene photos and videos. This act falls under cybercrime.

How are Deepfakes Dangerous?

Since, this technology can be used to defame or blackmail any person and popular celebrities, hence, misusing of deepfakes is absolutely dangerous even more than 100 times from fake news. Indian government is working on it to control this situation.

The misuse of deepfakes can cause immense damage to democracy too. Just imagine, during an election, a video of a candidate is released in which he is making hate speech, racial remarks or any wrong statement against minority communities, thus the election process can be easily influenced.

However, some people think that deepfakes are just a tool and if it can be used positively like other tools then it will be beneficial to the society.

How to Detect Deepfakes?

There are some points which can be used to detect deepfakes –

• Take note of the expression. Facial transformations are almost always used in high-end Deepfake manipulations.
• Pay close attention to the cheeks and the brow. Is the skin extremely smooth or extremely wrinkly ? Is the skin’s aging similar to the aging of the hair and eyes? Deepfakes may be out of sync in some dimensions.
• Take note of the eyes and brows. Do shadows appear in the expected places? DeepFakes may not fully represent a scene’s natural physics.
• Take note of the glasses. Is there any reflection? Is there an excessive amount of glare? When the person moves, does the angle of the glare change? DeepFakes may once again fail to fully represent the natural physics of lighting.
• Take note of the facial hair, or lack thereof. Is this genuine facial hair? DeepFakes can change their mustache, sideburns, or beard. However, DeepFakes may fall short of producing completely natural-looking facial hair transformations.
• Take note of any facial moles. Is the mole genuine?
• Keep an eye out for blinking. Is the person blinking enough or blinking too much?
• Take note of the lip movements. Lip syncing is used in some deepfakes. Are the lip movements natural?

Is there Software to Detect Deepfakes?

Deepfakes can also be detected by some special softwares. Here are some examples –

• Sentinel
• Intel’s Real-time Deepfake Detector
• WeVerify
• Microsoft’s Video Authenticator Tool, etc.

Conclusion:

Thus, my friends, we have seen in the above article that deepfake is a tool, when misused, the society can face a lot of problems and even democracy can be in danger. The misuse of deepfakes is a cybercrime and hence the government should take appropriate action which it is doing.

That’s all for now my friends! I hope, you will find this article useful. I will always love your valuable feedbacks and comments. You can reach me through the email ids given below –

Yours

Abhijit Ranjan

bloggerabhi15@gmail.com

abhijitranjan15@gmail.com