Deepfakes are morphed online images that look real, often done with malicious intent. It’s a representation of a misleading incident. Some fakes are even pornographic, and use AI to represent people at events they were never at, say things they never did say.
What is a deepfake
Deepfake technology can create made-up photos, mimic people on the phone and do a lot of damage. These “voice skins” or “voice clones” of public figures are used to scam and defraud the opposite person. They have also been described as synthetic media, ‘digitally manipulated to replace one person’s likeness convincingly with that of another’ and ‘deep generative methods manipulate facial appearance.’
Why the name deepfakes for these doctored images? That’s because powerful tech like machine learning and artificial intelligence has been utilised to create them. All this involves training neural network architectures, such as autoencoders or generative adversarial networks (GANs). Hence these fakes are called deepfakes.
The Rashmika Mandanna incident
Animal actress Rashmika Mandanna’s recently became the victim of a deepfake video — one where she was shown getting into an elevator in revealing, body-hugging gym wear. Apparently, that isn’t her but someone else. And, the image has been heavily morphed where Mandanna’s face has been put over someone else’s. The Delhi police’s Intelligence Fusion and Strategic Operations (IFSO) Unit has taken charge of the matter and has written to Meta for more information regarding who could have posted the deepfake. Said the DCP, “We have written to Meta to access the URL ID of the account from which the video was generated. We have also started doing technical analysis.”
An FIR pertaining to the matter has also has been registered under sections 465 (forgery) and 469 (forgery for the purpose of harming reputation) of the Indian Penal Code and sections 66C and 66E of the Information Technology Act. The Delhi Commission for Women (DCW), after taking suo moto cognisance of the deepfake, also complained about it to the police. The case is under investigation with the police aiming to get to the bottom of things. It is said that the video was actually of British-Indian influencer Zara Patel who is known for her bold social media content. Rashmika’s face was morphed onto her face in the elevator incident.
What psychologists and IT experts say
Rumoured couple Shubman Gill and Sara Tendulkar’s morphed images are also doing the rounds and thought to be deepfakes. The fact that a deepfake looks so realistic is what makes it problematic. The Indian government has issued a strict reminder to social media platforms, telling them that impersonation online is illegal. As per reports, the Ministry of Electronics and Information Technology has told platforms to take down such content within 36 hours, a requirement mentioned in the IT Rules of 2021. It also urged for the exercise of due diligence and reasonable efforts to identify misinformation and deepfakes.
Shares psychotherapist Neeta V Shetty quite ironically, “There are no longer any deeper connections in society. Everything is social media centric. The people who create these deepfakes are simply trolls who want attention. We no longer pick up the phone and talk to people. Our lives revolve around social media and we depend on it for everything. Ethics have vanished from the society. We must understand that what people put on social media is not reality, it’s fake. We use those fake lives on social media as a standard to live our own lives. I’m not surprised by the existence of deepfakes. Trolls will go to any extent to malign someone’s personality, and cause harm to them. There is a degradation of values in society, and we don’t use social media mindfully. Such people who create deepfakes are bound to flourish in the kind of world we live in.”
Cyber expert and digital forensics trainer Anuraag Singh talks of deepfakes used for revenge. “People who make deepfakes are ill-motivated by various means such as greed, hatred, bad intentions, etc. They make deepfake AI videos to defame a person or take revenge from a person or organisation. To protect yourself from being a deepfake victim, one can follow social media privacy ethics. Laws have been made to bring down deepfake videos in 36 hours. One can also report deepfake crimes under section 66 C and sector 66 D of IT Act.”
Entering web 3.0
Cyber expert Mieet Shah’s company received around 600 deepfake complaints over the past month. “Deepfakes are based on AI. They are created by different hackers and memers. Many times, the aim is to blackmail people for money, by putting up naked images of them online. What one can do to protect oneself from such incidents is self copyright one’s own pics and prevent others from using them,” Shah says. “Also another thing one can do is add and follow only known people on social media, to avoid hackers from getting access to your account. Most people who create deepfakes don’t have any specific motive. It’s done mainly to gain popularity or to blackmail someone. Now we are moving to Web 3.0 and hence there are many opportunities for people use AI to create things like deepfakes.”
