EXPOSING THE DANGER OF DEEP FAKES: A LEGAL OVERVIEW

ABSTRACT:

AI-generated altered media, or “deep fakes,” are becoming a bigger menace to people, businesses, and society as a whole. This research paper offers a thorough legal study of the deep fake phenomena, emphasizing the dangers they pose for disinformation, security, and privacy. We review the current legal frameworks and suggest legislative and technical measures as well as other possible ways to deal with this changing situation. The research provides a comprehensive overview of the legal landscape around deep fakes by examining ethical issues, and the role of developing technology. Policymakers, lawyers, and technologists can fight the risks of deep fakes while protecting basic rights and maintaining public confidence in digital media by using this study, which is a valuable resource.

KEYWORDS: Deepfake, privacy laws, defamation, IPR, GANs, IPC.

INTRODUCTION:

“It has become appallingly obvious that our technology has exceeded our humanity.” – Albert Einstein[1]

 In this era of AI-driven deception, what you see may not be what you get. Have you ever seen a viral video where a girl who looks exactly like Rashmika Mandanna walks into an elevator casually, and it becomes viral, causing a lot of people to share and talk about it? Or maybe I saw the photo of Shubman Gill cozying up with Sara Tendulkar that caused a stir and sparked a plethora of assumptions? Let’s now move into another instance where the President of Ukraine, in a video, seems to say, “Lay down your arms and return to your families, Dying in this conflict is not worth it.” This video surfaced on the internet amid the Russia-Ukraine war[2]. Isn’t that concerning? Here’s the twist, though: it’s not true. Deepfakes have now emerged, an era when truthfulness is no longer visible.

 However, why “deepfake”? Like the video productions themselves, its response dances along the edge of being perceived. The name represents a fusion of two words: “Deep” which denotes deep learning as the artificially generated technology employed in creating such highly lifelike facades and “Fake,” which expresses its pretentiousness.

The confluence of technology with deception carries a terrible message because the lines are being blurred between what is original versus fake into a digital masquerade. Delving into the world of deepfakes unveils a labyrinth of legal complexities. Privacy laws find themselves trampled upon, identities pilfered, and reputations maliciously tarnished. The term “deep fakes” encapsulates a digital realm where reality is a puppet in the hands of illusionists.

RESEARCH METHODOLOGY:

The methodology adopted is mainly doctrinal in nature, drawing on primary and secondary data in a comprehensive manner, in order to gain insight into its object. A qualitative study design is adopted in order to examine the legal framework on deep fakes. Qualitative methods are appropriate for the study of deepfakes due to their multifaceted nature.

REVIEW OF THE LITERATURE:

Deep Fakes are a phenomenon that have been around for a while, but their popularity has recently grown over the internet. It’s critical to trace the history of deepfake back to 2014 and the advancements in artificial intelligence, especially with regard to GANs, which served as the foundation for the technology’s creation. The University of Montreal’s researcher, Ian Goodfellow invented GAN back in 2014 that revolutionized deep learning. A GAN consists of two components — generating and discriminating neural networks, which together result in high-quality pictures or films. The generator creates a counterfeit material which is then evaluated for its authenticity or validity by the discriminator. This process continues until it becomes impossible to distinguish between created and original materials. This is how GANs became the heart of deepfake technology.

In 2017, a Reddit user by the name of “deep fakes” gained some attention after he started posting manipulated videos of popular celebrities. Using the available open-source deep learning libraries and a big image dataset, the user synthesised seemingly realistic but imperfect deep fakes. Soon, other users created their own deep fakes that were also circulated on the Internet. The development of deepfakes occurred at a very fast pace, which gave rise to doubts regarding its possible abuse, including the distribution of non-consensual pornographic materials and alterations in the appearance of politicians.[3]

 In 2018, the term “deep fake” gained popularity as a result of an app that used deep learning methods to quickly replace faces in films. In its progression, deepfake technology overtook face-swapping applications. the dissemination of very realistic deepfake content intended for amusement or as a foundation for evil. The spike resulted in growing skepticism about its impact on misleading information, personal privacy, and public trust.

The issue reemerged recently with a video involving an Indian actress named Rashmika Mandanna going viral in the social media channels. This case brings to light the continuous complexities and outcomes that will be experienced due to the unrestricted distribution of fake deep content.

WHAT ARE THE LEGAL CHALLENGES FACED BY DEEP FAKES?

Deepfake technology’s emergence has resulted in a number of legal difficulties, including problems with defamation, intellectual property, privacy, and the possibility of abuse.

  1. PRIVACY LAWS:

-There are multiple issues that confront privacy laws when it comes to deep fakes. With the increasing sophistication in deepfake technology, privacy violations have become a primary issue.

-There are serious legal issues pertaining to copyright violation, including that concerning consent and ownership of one’s images in relation to deepfakes content.

-The quick development of deepfake abilities may outpace existing privacy legislation, resulting in legislative loopholes. Deepfakes use a person’s identity, face, or physical characteristics. This results in the right to privacy guaranteed by -Article 21 of the Indian Constitution being violated and abused. not just in terms of geographical and physical privacy, but also in terms of the transfer, storage, and safeguarding of personal data.

-The judiciary has demonstrated an attentive, inclusive attitude throughout the years via a number of rulings, acknowledging, defending, and upholding the right to privacy as an essential component of fundamental rights in the democratic state[4].

-Deepfake videos that are created and disseminated covertly, with no prior notification to the subject, highlights the significance of strong privacy protection. Invasive forms of manipulation render individuals susceptible, hence making the situation complicated since they have little remedy in law.

-The increasing technological innovations for making deepfakes require frequent revision of existing legislation with respect to protecting people’s privacy at all times and in highly advanced environments with complex laws.

  • INTELLECTUAL PROPERTY LAWS:

– Deepfake technology is a recent development and it has triggered complicated intellectual property issues requiring multiple solutions, and existing laws can’t effectively deal with this situation. The issues that deep fakes raise include those of unfair exploitation of image rights as well as copyright issues.

-Deep Fakes are mostly created using copyrighted images without permission thereby posing a threat to Intellectual property rights. People can have their likes used in many different ways without their permission, while the originators of the original content run a higher risk of their works being reused in deceptive or even malicious manner.

-Deepfakes blur the lines between real content, authentic content and manipulated content posing a challenge with enforcements of intellectual property law. The response entails a deliberate legal approach with focus on evolutionary characteristics of deepfake technology, copyright, and their implications in the changing world order. Navigating this complex legal terrain requires striking a balance between artistic expression, technological innovation, and protection of intellectual property rights.[5]

  • DEFAMATION:

-Recent development in defamation and deep fake technology has generated uncharted territories for lawsuits. With deep fakes, we are talking about overly realistic audio/video tampering capable of creating lies and presenting people in an unreliable fashion that might destroy their reputations. Faking such videos and photographs can generate wrongful and libelous assertions against the subject which cause difficulties in claiming liability against liable persons. Establishing who created a defamatory deep fake and why they did so is particularly difficult in court cases.

-Although its complexity is making traditional defamation laws struggle, it raises issues of liability, ethics and responsibility. With the deceitful element of deep fakes, the risk involved with this kind of damage increases and underscores the importance of legislation that is fit for such a complex situation as this one. It will be necessary to strike a balance between ensuring protection of people from harmful lies manufactured with utmost cunning and safeguarding freedom of speech while addressing defamation in the era where deep fakes reign.

  • NATIONAL SECURITY CONCERN[6]:

-Deepfake proliferation has raised the alarm about the growing threat of cyber attacks and espionage, taking the digital dangers to a higher level. Given their abilities to alter audio and visual data in a convincing manner, the deep fakes hold risks to informational and debatable material genuineness. One issue that arises relates to the abuse of deep fakes in political agendas by using fake created materials that may distort voters’ opinions on certain key issues and therefore compromise democracy.

The issue lies on the possibility that there could be a deepfake misuse aimed at distributing false information as well as fueling confusion especially for purposes of threatening national security. The use of technology for such purposes as imitating politicians and forgery of public statements is just one example that complicates the task of verifying information in modern time. There is a need for holistic approaches in development of national security strategies to combat deep fake threats which may potentially lead to undermining democratic institutions and disrupting national policies as well geopolitical relations.

  • FRAUD AND MISUSE:

– Deepfakes present a complex dilemma that encompasses issues of fraud, abuse, and challenges faced by consumers as well as companies. The most frightening aspect of deepfakes is that they can be employed by fraudsters as a means to commit different types of crimes.

-Deepfakes in impersonation may lead to financial fraud as well as social engineering among others. Added to these are possibilities of creating fake content with public figures and critical employees, worsening malevolent misinformation and manipulations more.

-Deepfake’s deceptiveness complicates the distinction between real versus unreal images, boosting chances people may be duped by a scam or trickery campaign. To combat fraud and misuse related to deep fakes, we need a holistic approach involving technical solutions for detection, strong legislation addressing fraudulent acts, and empowerment of people and organizations to recognize and prevent threats associated with these advanced forms of digital deceit.

WHAT IS THE ROLE OF TECHNOLOGY IN PROMOTING AND SPREADING OF DEEPFAKES?

Technology is key to both deepfake promotion and distribution, enabling the fast expansion of this phenomenon across all countries and spheres of life. Several key aspects highlight the influence of technology in the creation and proliferation of deep fakes:

1. Advanced Deep Learning Algorithms: – GANs are one of them and they got very advanced lately. Such algorithms provide for realistic artificial (audio, visual) generation, which complicates the distinction between authentic and artificial (audio, visual) media.[7]

2. Accessibility of Tools and Software: – Deepfake creation is a process that has been made readily accessible to users. Deepfake platforms such as FakeApp among many others have provided means of production such that even with low-level expert knowledge a person can produce realistic deep fake content.

 3. Increased Processing Power: – As a result, the improvement in computational ability and GPUs makes it possible for creating deep fakes with more accuracy and at an increased rate. This facilitates immediate creation of good stuff as it has promoted vast manufacturing of deeper fakes in society.

4. Social Media Platforms and Viral Spread: – Deepfakes can travel fast via technology using social-media. Deepfake material spreads through other websites such as facebook, twitter, and youtube because it is viral in nature.

 5. Automated Content Generation: – Deep fake software used in producing manipulated media in great amounts. These allow for easy production and distribution of deep fakes which present challenges for social media platforms in their detection and control.

 6. Realistic Facial and Voice Synthesis: – Moreover, recent developments in facial recognition and voice synthesis make deepfakes appear real. The realism in manipulated content increases its deceitfulness, undermining a person’s ability to distinguish true media from false ones.

WHAT ARE THE MAIN TOPICS OF CONCERN UNDER DEEPFAKES?

1. Forensic Issues[8]:

  • Deep Fakes pose major problems in forensic situations, where accurate evidence is imperative. However, deep fakes can be used to manipulate audio and visual recordings, challenging the credibility of digital evidence during court case processes. For instance, what if a legal case were based on a deep fake video?
  • Provision of powerful forensic approaches in identification and confirmation on the genuineness of digital material stresses the possibility of mislead jury or judge. The rapidly proliferating deepfake technology presents challenges for maintaining the validity of evidence in legal investigations through forensic specialists.

 2. Pornography:

  • The integration of deep fakes, in the context of pornography raises legal concerns. Both famous individuals and ordinary citizens face the risk of having their images manipulated and used in content without their consent. This invasion of privacy not only affects them but also has an impact on their children causing potential reputational harm and psychological consequences.
  • Disturbingly there have been cases where non consenting individuals were featured in deepfake films resulting in harm and the unauthorized distribution of explicit material. As deep fake pornography continues to proliferate, women are at the forefront of fighting against the use of intelligence to impose innocent peoples faces onto explicit and demeaning content without their knowledge or consent. This horrifying violation subjects victims to humiliation and trauma. However it is crucial to move beyond these stories to fully comprehend the gravity of this issue.
  •  Addressing the legal challenges posed by production of deepfake pornography requires a comprehensive approach that encompasses legislation, technological defenses and support systems for victims[9].

3. LAW:

  • Law surrounding deep fakes is a very complicated matter. Intellectual property, privacy, defamation, liability of the platforms hosting the content of deep fakes increase the complexity to these legal questions. For instance, it remains hard to make people answerable for sharing harmful deep fakes whenever social networking websites are held liable for dispensing such ones.
  • A case in point is where a deep fake depicting an influential personality using vulgar words created conflicts on the host’s accountability. Therefore, it is important for the government to make laws that ensure there is a balance between free speech and avoiding harm among the citizens in the wake of deep fake technology.

HOW DIFFERENT LEGAL JURISDICTIONS APPROACH DEEPFAKE?

  • China:
  • Enacted a 2019 law requiring disclaimers for individuals and companies using deepfake technologies.
  • Illegal to disseminate deep fakes without explicit disclosure of artificial synthesis.
  • Cyberspace Administration of China (CAC) introduced comprehensive laws for deepfake suppliers[10].
  • United Kingdom:
  • Invested in research and collaborated with academics and businesses to develop deep fake recognition protocols.
  • Funded research and development through the ENOUGH communications campaign to address revenge and deep fake porn.
  • Upcoming Online Safety Bill in the UK to address the production and dissemination of malicious deepfakes.
  • South Korea:
  • Passed legislation in 2020 criminalizing the distribution of deep-fake content.
  • Advocates for additional measures, including education, civil remedies, and recourse against digital pornography and sex crimes.
  • Canada:
  • Implemented a three-pillar anti-deepfake action plan: prevention, detection, and response.
  • Focus on increasing public awareness, developing prevention technologies, and investing in deep fake detection algorithms.
  • Proposed bill considering outlawing the production and distribution of deep fakes in bad faith.
  • Existing Canadian law prohibits the circulation of non-consensual intimate image disclosure.

ARE THERE ANY LAWS TO TACKLE THE DANGER OF DEEPFAKE IN INDIA?

  • In India when somebody becomes a victim of a deep fake crime such as taking pictures, publishing them in the media, or transmitting such information among people through a computer they are subject to Section 66E under the Indian Information Technology Act of 2000. For this offense, the maximum punishment could be a fine of two lacs rupee or three years imprisonment.
  • Another significant section is Section 66D of the IT Act. It means that any person who intentionally falsifies computer resources or communication resources for the purpose of deceiving or impersonating another is liable to be imprisoned for three years or fined up to ₹1 lakh. These laws included in the Indian IT Act are applicable to the prosecution of criminals involved in deep fake cyber crimes.
  • Under Section 124A of The Indian Penal Code, 1860, an individual or group shall be deemed to have committed the offense of sedition if they endeavor or are successful in using deepfakes to stir up hatred and anti-national sentiments against the rule of law, the current government, or to bring contempt against the nation and its democratic principles.
  • Sections 67A and 66 B of the Information Technology Act of 2000 apply to deep fakes that include pornographic material. The parts describe a suitable fine for publishing and disseminating graphic and sexual content involving adults and minors. As a result, some websites, including Pornhub, have outlawed graphic sexual content.
  • Deep Fakes are not covered by a specific law in India though there are some legal provisions and government’s initiatives that may be applied towards solving the problem. Indian Laws do provide a block of laws that challenge deepfakes indirectly but given the increasing sophistication of deep fakes, the Indian government may have no choice but to come up with stronger measures aimed at ensuring citizens are not harmed during this time.[11]

 SUGGESTIONS AND CONCLUSION:

Considering the subtle complexities behind the deepfake dilemma reinforces the necessity of working together in advance in order to avoid possible obstacles. In national terms, there is a loud call for having carefully designed and well stipulated legal frameworks geared towards combating the complex intrigues associated with the field of deepfake technology. It is not just a legal question, but also a technology one for the frontier. It is not only about developing new detection means but also applying blockchain in order to protect ourselves from emerging tricked manners. However, understanding of the public lies at the core of the matter rather than just lawful advances and innovations. Such initiatives like media literacy programs or the implementation of ethical use guidelines may help people separate the useful wheat from the useless chaff in the digital world. Privacy laws require a totally different thinking if it has to be implemented from suggestion. Therefore, stricter rules should be enacted with regard to matters such as consent and ownership of images. However, another vital aspect is that such provisions should be swift enough to counter the constant advancement of technology. At last, it is necessary to step back and look at the broader picture. It also means that international cybersecurity standards should be established in order to provide protection from deepfake-based cyber threats. In actual sense, it serves as a call for joint actions among lawmakers, IT specialists and the general population to reach a compromise. This concludes that this research seeks to add to the existing perspectives on this discussion. It is more than an observation; it is also a zealous call on all people to take a stand against the menace posed by deepfakes in this dynamic environment where technology is evolving.

 AMIT KUMAR PATRA

 NATIONAL LAW UNIVERSITY ODISHA.

REFERENCES:

  1. Telegraph Reporters, “Deepfake” Video Shows Volodymyr Zelensky Telling Ukrainians to Surrender, The Telegraph, Mar. 17, 2022, https://www.telegraph.co.uk/world-news/2022/03/17/deepfake-video-shows-volodymyr-zelensky-telling-ukrainians-surrender/ (last visited Nov 12, 2023).
  2. Marcin Frąckiewicz, The Evolution of Deepfake: Tracing the History and Development of AI-Generated Content, TS2 SPACE (Jul. 19, 2023), https://ts2.space/en/the-evolution-of-deepfake-tracing-the-history-and-development-of-ai-generated-content/ (last visited Nov 12, 2023).
  3. Rachit Garg, Is Indian Law Equipped Enough to Deal with Deepfakes, iPleaders (Nov. 11, 2021), https://blog.ipleaders.in/is-the-indian-law-equipped-enough-to-deal-with-deepfakes/ (last visited Nov 12, 2023).
  4. The Trademark Lawyer Magazine | Intellectual Property Specialist, Trademark Lawyer Magazine, https://trademarklawyermagazine.com/ (last visited Nov 12, 2023).
  5. The Uses and Abuses of Deepfake Technology, Canadian Global Affairs Institute, https://www.cgai.ca/the_uses_and_abuses_of_deepfake_technology (last visited Nov 13, 2023).
  6. Deepfake pornography could be a growing problem as AI editing programs become more sophisticated – CBS News, (2023), https://www.cbsnews.com/news/deepfake-porn-ai-technology/ (last visited Nov 13, 2023).
  7.  Fintelics, Deepfakes and AI: Understanding the Technology and Its Implications, Medium (May 14, 2023), https://fintelics.medium.com/deepfakes-and-ai-understanding-the-technology-and-its-implications-794dfa114ef5
  1. Amanda Lawson, A Look at Global Deepfake Regulation Approaches, RAI Institute (2023), https://www.responsible.ai/post/a-look-at-global-deepfake-regulation-approaches (last visited Nov 12, 2023).
  2. Dafeng Gong, Deepfake Forensics, an AI-Synthesized Detection with Deep Convolutional GenerativeAdversarial Networks, 9 IJATCSE 2861 (2020).
  3. Vasundhara Shankar, Deepfakes Call for Stronger Laws, BusinessLine (2023), https://www.thehindubusinessline.com/business-laws/deepfakes-call-for-stronger-laws/article67077019.ece (last visited Nov 11, 2023).
  4. Deepfake pornography could be a growing problem as AI editing programs become more sophisticated – CBS News, supra note 6.

[1] Ken Makovsky, Is Technology Exceeding Humanity?, Forbes, https://www.forbes.com/sites/kenmakovsky/2012/05/07/is-technology-exceeding-humanity/ (last visited Nov 28, 2023).

[2] Telegraph Reporters, “Deepfake” Video Shows Volodymyr Zelensky Telling Ukrainians to Surrender, The Telegraph, Mar. 17, 2022, https://www.telegraph.co.uk/world-news/2022/03/17/deepfake-video-shows-volodymyr-zelensky-telling-ukrainians-surrender/  (last visited Nov 12, 2023).

[3] Marcin Frąckiewicz, The Evolution of Deepfake: Tracing the History and Development of AI-Generated Content, TS2 SPACE (Jul. 19, 2023), https://ts2.space/en/the-evolution-of-deepfake-tracing-the-history-and-development-of-ai-generated-content/  (last visited Nov 12, 2023).

[4] Rachit Garg, Is Indian Law Equipped Enough to Deal with Deepfakes, iPleaders (Nov. 11, 2021), https://blog.ipleaders.in/is-the-indian-law-equipped-enough-to-deal-with-deepfakes/ (last visited Nov 12, 2023).

[5] The Trademark Lawyer Magazine | Intellectual Property Specialist, Trademark Lawyer Magazine, https://trademarklawyermagazine.com/ (last visited Nov 12, 2023).

[6] The Uses and Abuses of Deepfake Technology, Canadian Global Affairs Institute, https://www.cgai.ca/the_uses_and_abuses_of_deepfake_technology (last visited Nov 13, 2023).

[7] Fintelics, Deepfakes and AI: Understanding the Technology and Its Implications, Medium (May 14, 2023), https://fintelics.medium.com/deepfakes-and-ai-understanding-the-technology-and-its-implications-794dfa114ef5 (last visited Nov 19, 2023).

[8] Dafeng Gong, Deepfake Forensics, an AI-Synthesized Detection with Deep Convolutional Generative Adversarial Networks, 9 IJATCSE 2861 (2020).

[9] Deepfake pornography could be a growing problem as AI editing programs become more sophisticated – CBS News, (2023), https://www.cbsnews.com/news/deepfake-porn-ai-technology/ (last visited Nov 13, 2023).

[10] Amanda Lawson, A Look at Global Deepfake Regulation Approaches, RAI Institute (2023), https://www.responsible.ai/post/a-look-at-global-deepfake-regulation-approaches (last visited Nov 12, 2023).

[11] Vasundhara Shankar, Deepfakes Call for Stronger Laws, BusinessLine (2023), https://www.thehindubusinessline.com/business-laws/deepfakes-call-for-stronger-laws/article67077019.ece (last visited Nov 11, 2023).