ABSTRACT
Using only a series of images and videos of a person and publicly available software, it is now possible to insert the person’s likeness into a photo or a video and show them saying or doing almost anything that never actually happened. Deep learning has solved various problems effectively, from high data analytics to computer data to human level problems. However, has been used to produce software that interferes, threatens and hampers a person’s privacy, democracy and national security. This deep learning technology is widely known as “Deepfake Technology.” This technology might seem interesting but it could spread as misinformation via various social media platforms. Deepfakes contents could be dangerous for individuals as well as for our communities, organization etc. This research paper attempts to scrutinize this technology, it’s pros and cons and the people responsible. The scholar attempts to explain the deepfake technology, its impact, the existing laws for the same in India and outside India and the loopholes which the existing laws has, the proposed reforms and what are the suggestions to overcome from this technology threat.
KEYWORDS
Deepfake, GANs, Artificial intelligence, privacy laws, Legislation, IT Act
INTRODUCTION
The word “Deepfake” refers to audio and video files produced using adversarial deep learning methods, namely Generative Adversarial Techniques (GANs). To analyze data, GANs consist of a generator and discriminator. Using a thousand images of a subject, GAN may generate a new portrait that roughly resembles the originals without being an exact replica of any one of them. Since they employ actual footage, can contain real voice, video, and photographs, and spread swiftly on social media, deepfakes are hard to spot. Artificial Intelligence (AI), Photoshop, machine learning (ML), and numerous other tools are widely utilized in the production of deepfake films, clips, and other media. The unethical applications of Deepfake technology have negative long- and short-term effects on our society. However, there could be a lot of advantages to using this technology properly. Below is a detailed description of Deepfake technology’s both positive and bad applications.
Additionally, deepfake technology has beneficial applications in a wide range of industries, including as digital communications and instructional media, games and entertainment, social media and healthcare, and film.
There are several ways that deepfake technology might help the film business. It can be used to update film footage or create synthetic voices for performers who have lost theirs. Filmmakers will have the ability to remake old or new films with actors who have passed away. That is how the late actor Paul Walker’s resurrection for Fast & Furious 7 came to be. Deepfake technology also makes it possible to dub audio in any language, which lowers language barriers and improves the viewing experience for a wide range of audiences. Positive applications of technology can also be found in social media networks. Moreover, it can be used to digitally reconstruct an amputee’s limb or help transgender individuals better understand their desired gender. Sessions can be made more dynamic and interesting by utilizing deepfake technology. For better comprehension, it can be utilized, for instance, to generate historical events or people interacting with virtual characters.
Still, there are possible dangers connected to deepfake technology. It is able to disseminate misleading information. It might be used, for instance, to generate videos that give the impression that something happened or was done by someone, even when it never did.
Manoj Tiwari, the leader of the Bharatiya Janata Party (BJP), is purportedly heard speaking in Hindi, English, and Haryanvi in a deepfake video that was made in 2020. The movie was shared on a number of WhatsApp groups prior to the 2020 Delhi Legislative Assembly elections. Moreover, deepfake technology can be used to harm someone’s reputation by, for instance, fabricating a video showing a celebrity acting in a way that they have never done and never will. Actor Rashmika Mandanna was featured in a video that went viral on social media on November 7, 2023, shocking many online. The brief footage, which included Mandanna’s likeness, showed a lady wearing a bodysuit entering an elevator. Using deepfake technology, the original video, which featured British Indian influencer Zara Patel, was altered. According to reports, the Delhi Police Special Cell has filed a formal complaint against unidentified individuals under Sections 465 (forgery) and 469 (forgery intended to do harm to the reputation of a party). Actress Rashmika came out and wrote about the genuineness of an unchaste video that had appeared on numerous social media sites the day before. “Anything like this is honestly, really scary—not just for me, but for all of us who are vulnerable to so much harm today because of how technology is being misused,” the author said.
RESEARCH METHODOLOGY
This paper is based on secondary sources for the deep analysis of the technology called deepfake, it’s pros and cons, the existing laws in India and outside India. Secondary sources of information like newspapers, journals and websites are used for the research.
REVIEW OF LITERATURE
Mika Westerlund’s book, The Emergence of Deepfake Technology: A Review, provides a basic introduction to deepfake technology. It describes how machine learning methods, deepfakes, and Generative Adversarial Networks (GANs) can be applied for both evil and humorous purposes.
In “Deepfakes Call for Stronger Laws,” a news article from The Hindu Business Line, it is discussed how urgently stronger legal frameworks are needed to handle the problems caused by deepfakes. It draws attention to the legal ramifications of deepfake technology, especially in India. In order to safeguard people from hazards like identity theft and character assassination, the article highlights the potential for abuse in these areas and the significance of strict rules.
The SCC Online Blog: “Legal Status of Handling Crimes Associated with Deepfakes in India” looks at the legal framework for handling crimes using deepfakes, with an emphasis on the Indian legal system. It examines the challenges Indian law has in keeping up with emerging technological advancements. The blog highlights how adaptable the legal system needs to be in order to handle these contemporary technology concerns and investigates potential legal remedies for deepfake offenses.
An SSRN research paper titled “Deepfakes and the Law: Legal Implications of Deepfake Technology” is available. This research paper presents an academic analysis of the legal implications of deepfake technology. It addresses subjects like intellectual property difficulties, privacy infringement, and defamation. In order to successfully handle the concerns associated with deepfakes, the paper proposes for revisions to the law and underlines the necessity of comprehensive legal frameworks.
When all is said and done, these resources aid in giving a complete understanding of deepfake technology and its legal ramifications. From general overviews to in-depth legal studies, the literature highlights the need for changing legal frameworks to address the concerns brought up by deepfakes and protect individuals and society from the risks related with their exploitation. The knowledge gained from these diverse sources will be useful in understanding the legal ramifications of deepfake technologies.
DEEPFAKE AND ITS IMPACT
Deepfakes are the product of Artificial Intelligence (AI) that merge, combine, replace images and video clips to create fake videos that appear authentic. Although they have clear benefits in many sectors, they can also be used to exploit people, sabotage elections and spread information worldwide. Editing tools like Photoshop have been used for decades, the first ever use of deepfake technology can reportedly be traced back to Reddit contributor who in 2017 had used a publicly available AI-driven software to create pornographic and sexual content by using the faces of celebrities on the bodies of ordinary people.
Deepfakes can also lead to traumas especially for women and can also bring suicidal thoughts if someone’s reputation is lost which they created during their lifetime. Such thoughts can also come in the minds of a middle-class people especially if the reputation of their girl child is hampered or lost. “The tools to create and disseminate disinformation are easier, faster, cheaper and more accessible than ever”, reported by the Deeptrust Alliance in the year 2020.
As deepfakes and other similar technologies become harder to detect, more resources are now accessible to protect individuals against such misuse. For instance, the Massachusetts Institute of Technology (MIT) created a Detect Fakes Website to help people identify deepfakes.
LAWS REGULATING DEEPFAKES IN INDIA
The Information Technology Act, 2000
It also reaffirmed that organizations risk losing the protection provided by Section 79(1) of the IT Act if they do not comply with the pertinent provisions of the Information Technology Act, 2000 (IT ACT) and Rules, Rule 7 of the Information Technology Rules Code, 2021 (hereinafter referred to as the “IT Rules”). Online intermediaries are released from liability for any third-party data, information, or communication link that they make available under this section of the IT Act. The Indian Penal Code gives victims the ability to bring platforms before courts under Rule 7 of the IT Rules.
Sections 509 (words or acts designed to offend a woman’s modesty), Section 499 (criminal defamation), and Section 153 (a) and (b) (spreading hate on communal lines) of the Indian Penal Code, 1860, IPC can also be revived for cybercrimes related to deepfakes.
A person who violates the privacy of another by posting or transmitting an image of their private area without that person’s agreement faces up to three years in prison and a fine of INR two lakh, according to Section 66E of the IT Act.
The IT Act expressly forbids and stipulates penalties for publishing or sending pornographic or electronically reproduced sexually explicit information in sections 67, 67A, and 67B, respectively. Urce uses impersonation to cheat. According to Section 66D of the IT Act, anyone caught using a communication device or computer resource to deceive someone by impersonating them faces a three-year sentence and a fine of up to one lakh rupees.
If any video or image that is protected by copyright has been used to produce deepfakes, the Copyright Act of 1957 may be invoked. Any property that belongs to another person and over which they have the sole right to use is prohibited by Section 51.
IS THERE ANY LACUNA IN THE EXISTING LAWS?
Since that the regulations that are in place were not developed with emerging technologies like deepfake in mind, they are really insufficient. Shehnaz Ahmed, head of fintech at Delhi’s Vidhi Centre for Legal Policy. She did, however, note that proposing changes to the law is not the answer. According to her, there is a current moral panic stemming from high-profile incidents, but the main subject that needs to be addressed is how India should regulate future technology like artificial intelligence.
She emphasizes that the foundation of such a framework needs to be a market analysis of the various forms of harm caused by AI technology. She draws attention to a gap in the current IT Rules, stating that preventive measures have not been prioritized and that the rules mainly address situations in which illicit content has been posted and causes harm. Although Mr. Gupta acknowledges that the current rules need to be changed, he also notes that these laws primarily address online censorship and do not take into account the advantages or workings of artificial intelligence.
The victim bears the responsibility for filing a complaint under the statutes. He claims that many of them have had unsatisfactory experiences with the local police departments in terms of their investigations or the wrongdoer receiving any sort of punishment.
INTERNATIONAL LAWS
President Joe Biden of the United States issued an executive order in October 2023 to address the threats linked with artificial intelligence (AI), which ranged from privacy concerns to national security issues. The memo charged the Department of Commerce with creating guidelines for “watermarking,” or the process of tagging content created by artificial intelligence. Meanwhile, a few jurisdictions, including Texas and California, passed laws making it illegal to produce and distribute deepfake films. Interestingly, a law against the distribution of deepfake pornography was passed in Virginia.
Congress introduced the DEEP FAKES Accountability Bill in 2023 in reaction to these developments. The bill requires creators to identify deepfakes on online platforms and alert consumers to any changes made to videos or other content. In addition, the measure suggests that those who create and distribute malevolent deepfakes face criminal consequences.
Globally, the European Union (EU) strengthened its Disinformation Code of Practice to force major social media companies like Google, X, and Meta to install deepfake content detection systems or risk paying hefty fines. The Digital Services Act has given the Code of Practice, which was first intended to be a self-regulating tool, more regulatory teeth.
The aforementioned legislative and regulatory measures indicate a deliberate endeavour, both nationally and globally, to tackle the obstacles presented by deepfake technology. Policymakers are actively involved in developing the legal framework surrounding AI-generated content, with goals ranging from bolstering national security to safeguarding personal privacy and battling misinformation.
LAWSUITS THAT PAVED THE WAY AGAINST DEEPFAKES
According to a ToI report, on November 30, 2023, thieves used a retired IPS officer’s video to extort a 76-year-old man in the UP Police. Fearing that he would be sued by the authorities for what appeared to be sex solicitation, the elderly man repeatedly paid the offenders.
A committed team has been assembled to investigate, and a formal formal complaint has been filed. Ghaziabad resident Arvind Sharma received his first smartphone and created a Facebook account. They first communicated via Facebook video call on November 4, 2023. When he saw a picture of someone in the nude, he immediately cut the, but the criminals had enough time to trap him. A little later, he got a video call on WhatsApp, and this time, the caller was threatening him in a police uniform. The individual in the video threatened to disclose a phony video of Sharma to his family and file a complaint against Sharma’s father if he was not paid. Sharma transferred Rs. 5000 out of humiliation. He paid more and more out of a constant fear of disgrace, totalling Rs. 74,000. He even needed to borrow money from his business.
After a few days, the offender sent in additional demands for money, and the man committed suicide as a result of the pressure. When he finally brought his family in, they recognized the footage that had former ADG Prem Prakash in it. Next, they went up to the Ghaziabad police. The cyber cell assisted in the investigation, and the case was filed under the IT Act. The police added that in order to get further information about the criminals’ accounts, they will be writing to Facebook, WhatsApp’s parent company, Meta, and Facebook.
X, Meta is asked by the Ministry of Electronics & IT to eliminate deepfakes created by AI ahead of the Lok Sabha elections. Deepfakes are artificial intelligence-generated videos that mimic real people using their photos or videos. This technology’s potential for misuse and disinformation has made it a major global concern. The Ministry has directed them to abide by IT Rules 2021 and has asked for the proper usage of AI. Additionally, Prime Minister Narendra Modi expressed his worry about the improper application of deepfake technology.
Kerela got its first report on deepfake in July of 2022. Radhakrishnan, the 73-year-old victim, received a WhatsApp call from Venu Kumar, his coworker. He immediately transferred the funds because he thought his colleague was in urgent need of money. The caller, who used deepfake technology to perfectly replicate his colleague’s voice, begged for a loan of $40,000. Through his social media accounts, the scammer obtained his personal information. As a result of the case, the authorities were able to link the funds to a Maharashtra account and released a warning regarding deepfake schemes.
Anil Kapoor v. Simply Life India and Ors.
The actor launched a lawsuit after discovering his AI-generated video. The Delhi High Court has protected actors’ personas and personal characteristics from being misused, particularly when it comes to the usage of AI techniques to create deepfakes. The actor’s name, likeness, and image were effectively prohibited from being used for financial benefit or commercial purposes by sixteen (16) businesses by the Court’s ex- parte order, which also prohibited them from using technical tools like artificial intelligence.
Amitabh Bachchan v. Rajat Negi and Ors.
Ad interim in rem injunction was issued to the renowned actor to prevent the use of his image, voice, and other personal characteristics for commercial purposes.
SUGGESTIONS
The aggrieved person should file First Information Reports (FIRs) at their nearest police station and ask for the remedies available under the various provisions which cover the Deepfake technology or cyber security. Pictures and videos are sensitive content of an individual as defined under the Digital Personal Data Protection Act, 2023. But this Act itself has many loopholes which needs to be filled or addressed. Deepfakes are thus violation of the right of privacy under Article 21 of an individual. Data which are available publicly will not fall under the law. But the data used without the consent of the concerned party which he/she does not put it publicly on social media platforms definitely falls under the law. The social media owners or the parties concerned will have to own up if the information put on their sites or any social media handle can be mined for the purpose of spreading misinformation.
CONCLUSIONS
The negative effects of deepfakes will keep politicians from putting out new legislation. Though Deepfakes technology is very new, lawmakers should take this into account when contemplating such laws as it does not present a unique set of issues. In addition to possible unintended consequences like the entrenchment of market incumbents, attempts to address Deepfake political meddling and altered pornography should take into account the potential impact such suggestions may have on First Amendment-protected speech. Laws aimed at combating Deepfakes ought to be restricted to a single type of content that aims to cause particular harms.
REFERENCES
[1] Ashraf A. Abu-Ein, Obaida M. Al-Hazaimeh, Alaa M. Dawood, Andraws I. Swidan, “Analysis of the current state of deepfake techniques-creation and detection methods,” Vol. 28, No. 3, Indonesian Journal of Electrical Engineering and Computer Science, 1660-1662 (2022)
[1] MikaWesterlund, “The Emergence of Deepfake Technology: A Review”, Vol. 9, Issue 11, Page 41-45 (2019)
[1] Deepfakes And Breach Of Personal Data – A Bigger Picture (livelaw.in)
[1] https://cybercert.in/what-is-deep-fake-cyber-crime-what-does-indian-law-say-about-it/
[1] Bahar Uddin Mahmud, Afsana Sharmin, “Deep Insights of Deepfake Technology: A Review” Chittagong University of Engineering & Technology, Bangladesh
[1] The Information Technology Act, 2000, S.66, S.77, S.79, No.21, Acts of Parliament,2000 (India)
[1] https://www.thehindu.com/news/national/regulating-deepfakes-generative-ai-in-india-explained/article67591640.ece/amp/ (Last visited on 12.03,2024)
[1] https://www.thehindu.com/news/national/regulating-deepfakes-generative-ai-in-india-explained/article67591640.ece/amp/ (Last visited on 13.3.2024)
[1] https://m.economictimes.com/news/new-updates/man-gets-caught-in-deepfake-trap-almost-ends-life-among-first-such-cases-in-india/articleshow/105611955.cms (Last visited on 12.03.2024)
[1] https://economictimes.indiatimes.com/topic/deepfake (Last visited on 11.03.2024)
[1] https://www.indiancybersquad.org/post/case-study-kerala-s-first-deepfake-fraud (Last visited on 13.03.2024)
[1] https://www.livelaw.in/amp/law-firms/law-firm-articles-/deepfakes-personal-data-artificial-intelligence-machine-learning-ministry-of-electronics-and-information-technology-information-technology-act-242916 (Last visited on 12.03.2024)
[1] https://www.livelaw.in/amp/law-firms/law-firm-articles-/deepfakes-personal-data-artificial-intelligence-machine-learning-ministry-of-electronics-and-information-technology-information-technology-act-242916 (Last visited on 11.03.2024)
AUTHOR:
Khushi Bhagat
University of Engineering and Management, Kolkata