Some time ago, after a very objectionable clip linked to Prime Minister KP Sharma Oli was made viral through deepfake technology, CPN Unified Socialist leader and former minister Ram Kumari Jhankri became angry. She, who is active on social media and media, wrote on Facebook, "One of my friends said: So many cyber crimes are being committed against you, why are you silent?" Either don't present yourself on social media or take action to bring such people to the legal sphere. You also tell me what should I do?'

His allusion was not only to those using deepfake to dig him up, but also to the vulgar and profanity used in commenting on his posts on social media. By the time this article was prepared, his post had received 883 suggestions. Most of them have advised to take legal action against those who make deepfakes, not tolerating online abuse. Jhankri said that she has been living well so far and is taking advice on what can be done about it. 

This is just one representative case of politicians like him or other celebrities and public figures, using artificial intelligence (AI) and even deepfakes. Kathmandu Metropolitan City Deputy Mayor Sunita Dangol filed a complaint with the Cyber ​​Bureau of Nepal Police after a similar indecent deepfake clip with Mayor Balendra Shah went viral. After his complaint, the police arrested a young man who lives in Bajura and runs a medical business in Sindhupalchowk for making such content. According to the police, deepfakes like those of other political figures were found on the young man's device. 

The singer review officer had also filed a complaint with the Cyber ​​Bureau in December last year, saying that she had misused the photo she posted on social media and created an objectionable deepfake.

She had mentioned in the petition that she was suffering from mental stress, saying that an offensive video was posted on a Facebook page called 'Dipu Kancho'. She wrote on Facebook, "Such content that affects my personal and family life has a negative impact on the society and the nation as well." The Cyber ​​Bureau has informed that people who misuse computers and AI will be investigated and action will be taken soon. For the purpose of cheap popularity, I request you not to do illegal activities that will negatively affect the life of any girl.'

Apart from these representative cases that have reached the scope of the law, deepfake content has also been created for the purpose of spreading misleading information, advertising, harassing someone, hashimjak or promoting a political agenda. In this, there is a trend among some internet users of today called the 'meme generation' to create deepfake videos, audios and images for both positive and negative purposes. Deepfakes were popularized in 2017 by a Reddit user who hacked celebrity photos. However, since the 1990s, the deepfake style has been used in a variety of ways to manipulate images. 

Ever since the use of AI became widespread in 2020, deepfakes have spread to audio, gaming, virtual reality, writing, and educational and training purposes along with photos and videos. Deepfake has also been used in positive ways such as making movies more artistic, improving the gaming experience, making virtual reality more alive, training or making the field of comedy more effective. However, it seems that its negative use has started to make a difference in society and information flow and in some cases even in political results. 

The god who gets up and walks because of AI

The use of deepfakes is not limited to politics and celebrities. Lord Vishnu of the Budhanilkanth temple complex, who has been lying comfortably in the water for centuries, now suddenly opens his eyes, gets up and starts walking. The great Kailashnath Mahadev of Sanga has left his throne and descended the ladder to meet his devotees. Meanwhile, even the Buddhist stupa is not quiet, and the top canopy of the stupa is moving and the statue of Buddha appears inside the stupa.

Gods such as Bhairav, Buddha, Mahadev, Parashuram and other gods have started to get up and walk, these videos are now widely seen on social media such as Tiktok, Facebook, and Instagram. Is it some magic? Or, a new form of divinity? Especially this is the 'result' of modern technology. Powerful AI tools are bringing gods, idols, historical figures and old photographs to life. Such usage not only entertains and engages viewers, but also becomes a means of generating views and 'attention' for creators.

The old videos of King Birendra and Rani Aishwarya, which were previously found in poor quality on television or YouTube, are now circulating in HD quality on social media. In them, they are seen smiling and gossiping, while somewhere else they are walking in royal attire, Virendra's family is also seen hugging each other. Prithvi Narayan Shah is even seen dancing in songs like 'Bajaer baja, lin ayo raja..'. Such videos are created using deepfake AI tools.

Nepali rupees, elephants, gardens, and zarayos are all running out of money and saying that 'the country is empty, now even the animals and birds will not live here', '20 rupees zarayos ran to the forest', 'New year did not come to me, like an elephant entered the forest,' Interesting captions like 'Goji Rittiyo' have been written. The five-rupee notes appear to be intertwined. Under the caption 'Effect of Earthquake', Dharhara is also seen on Tiktok in a funny clip of him walking gracefully on his robotic legs. AI-based deepfake technology risks spreading misinformation, breaching privacy and undermining credibility. In this way, some users are even surprised to see the 'animation' of gods, idols and deceased personalities.  How is

made deepfake? Before the use of

AI became widespread, the notion of deepfake was technically complex to embody. However, with the availability of machine learning and various online platforms and apps, many tools for creating such content have reached the hands of people. These platforms, available in both free and paid form, create deepfakes through the process of face swapping, matching and matching various related topics. These tools can be used to clone voices and make videos with almost matching movements and styles.

The use of deepfakes, which can be easily created as social network posts, messengers, fake screenshots of WhatsApp, obscene content, cartoons or animations intended to provoke, speech, and even dancing, is becoming widespread. Deepfake double videos in the voices of Prime Minister Oli, Congress President Sher Bahadur Deuba, Foreign Minister Arju Rana, National Independent Party's Ravi Lamichhane and others are scattered all over social media. 

AI apps like 'DeepFaceLab' create lifelike faces and animations based on old photos, pictures or videos of historical figures, idols or gods as mentioned above. Many tools like D-ID, Blender, Reface AI can be used to make pictures and images speak, walk, wink.

AI 'upscaling tools' are available on the Internet to convert old and historical photos and low-quality videos to high-quality ones. Such platforms have lifted not only the Dharahara of Kathmandu, but also the statue of Ballabhai Patel in India and the Statue of Liberty in America. Videos of Mahatma Gandhi groping to Indian rapper Honey Singh's songs at parties with six-pack abs are easy to find.

Last year, a deepfake video of popular South Indian actress Rashmika Mandana went viral. Those involved in the incident were also arrested by face swapping the video of another person who looked like him. The incident gained prominence in India's media debate about deepfakes. Later, the Government of India also made Rashmika a Special Envoy for Cyber ​​Security Promotion. Similarly, popular Indian actress Alia Bhatt became the victim of a deepfake made to match the trend of getting ready with an influencer, which is trending on Instagram and Tiktok. 

Around the same time, there was a legal dispute after a deepfake of famous Indian cricketer Sachin Tendulkar was found to be used to promote a controversial online gaming app. Using deepfake pictures, videos and voices of celebrities like Mahendra Singh Dhoni, Shah Rukh Khan, Amitabh Bachchan, Virat Kohli to fool people and confuse them with hashimjak is not a new thing in India. Ranveer Singh and Aamir Khan supporting any political party was a deepfake. Ranveer filed a police complaint about it. 

Not only that, in India, leaders like Narendra Modi, Rahul Gandhi, Mamata Banerjee, Arvind Kejriwal, Amit Shah, Nirmala Sitharaman, Kangana Ranaut have also become victims of deep fakes and 'manipulated content'. Indian Prime Minister Modi and Italian Prime Minister Giorgia Meloni have not only met, dated, married but also made vulgar content viral by social media and deepfake users. Similarly, Aam Aadmi Party chief Arvind Kejriwal becomes another star of Deepfake. It has not been long since Kejriwal's deepfake content went viral in association with Delhi's newly appointed Chief Minister Atisi. Deepfake speeches and pictures of Congress leader Rahul Gandhi, who is always targeted by the Bharatiya Janata Party's cyber bureau, are also being circulated. 

...

In India, which has nearly 600 million WhatsApp users and nearly a billion with internet access, some of these deepfakes are deliberately spread and widely reported in the media during last year's major elections. Not only negative, deepfake technology was also used to make DMK leader M Karunanidhi, who had passed away in the previous decade, among the workers, to address BJP candidate Manoj Tiwari in Delhi in different languages ​​and to make a video in memory of famous Indian artist Sushant Singh Rajput who died a few years ago. Such initiatives are highly appreciated.  The good use of deepfake made with

permission seems to be a helpful and accessible tool for reviving old music, improving gaming and virtual reality experience, making online classes and learning interesting, and spreading different languages. However, many people are suspicious of deepfake because the consequences of its misuse are extremely evil. Deepfakes have been widely used to spread their own propaganda in the Israel-Hamas conflict that has been going on for almost a year and a half. During the American presidential election, the things that both candidates Donald Trump and Kamala Harris did not say, did not say, and did not do were made into election campaign material through deep fakes and 'content manipulation'. 

The warring Russian and Ukrainian presidents Putin and Zelensky, the world's richest man Elon Musk, actor Tom Cruise, singers Taylor Swift and Beyoncé, football player Cristiano Ronaldo, along with American President Joe Biden, have been forced to endure the stress of deepfake audio, video and images. . Although such technologies help in entertainment, education and reinvention of history, there are many examples of its misuse causing problems. 

The worst example of deepfake use in the past year has been the platforms and Telegram bots used to portray various celebrities, particularly women, in nude or offensive situations. Such deepfake platforms and bots, which are easily and freely available by uploading someone's photo, have stressed both celebrities and ordinary people.

Even in Nepal, there have been complaints to the police about some such incidents of using deepfake to make obscene videos of Facebook, Messenger and other app users and blackmail them with the threat of making them public. Incidents and complaints of 'revenge porn', the use of AI and deepfakes to spread porn with the intention of causing pain after problems between people in a relationship or close relationship, have both increased.

The majority of the Internet is not well versed in digital literacy and the legal pitfalls that can arise from the use or dissemination of such content. For this reason, fake content is spread intentionally or even to collect views. Even the parties that seem indifferent to the regulation of AI 'generated and deepfake' content have started to fall victim to it in recent days. Deepfakes have emerged as another way to spread false and misleading information after the expression that the leader of the

party did not give, photos that were not taken, videos that were not known, statements or information that were not released, but also fake social network screenshots and messages with their own explanations. Such content is created not only about leaders and politics but also to increase tension communally, religiously, racially, sexually, regionally. As the use of deepfake with the intention of killing someone's character is increasing, the demand to regulate it has started to grow strongly.

Declare deepfake as a crime

Since 2017, Britain has prepared to put it on the list of criminal activities, saying that since 2017, revenge porn based on AI and deepfake, i.e. the act of making someone's vandalized obscene photos or videos, has increased by 700 percent. Britain is preparing to bring both those who make content and those who spread it into the scope of punishment for such acts targeting women and especially young girls. 

China has long been tough on deepfakes. In 2019, China introduced rules to clearly disclose deepfakes or content created with the use of AI, obtain permission, put a watermark, and make a complaint if someone makes an illegal deepfake. Deepfake regulation in the US is more focused on elections and voters, and at the same time, the US Parliament and the state of California have been trying to regulate the use of deepfake. 

The European Union has also added various provisions to its Digital Services Act and General Data Protection Regulation to require platforms to remove deepfake or AI 'generated content' that harms individuals or communities. The EU has repeatedly tried to improve the protection of unauthorized deepfakes and personal data. Similarly, South Korea has brought legal action against pornographic content created and disseminated without permission through the use of DeepFake. Canada has been working to raise awareness about deepfakes and AI 'generated content'. 

Similarly, Canada has begun efforts to regulate the problem of deepfakes by adopting a strategy of prevention, detection and response. Neighboring India has also laid the foundation for legal regulation of deepfakes by adding or amending clear provisions in laws such as Information Technology Act, Indian Penal Code, Data Protection etc. It seems that the government of India is trying to regulate fraud based on deepfake, production and dissemination of obscene content, defamation and various actions related to women's safety.