INTRODUCTION
Artificial Intelligence (A.I.) has evolved to the point where it can be used for everyday tasks; it can help you in every possible way you can think of. Now A.I. can generate highly realistic images, videos, & audios. In short, it’s a powerful tool that can be used to do something we never said or did before because of the limitations of the tools we had before (for the use of common people).[1]
While the technology has legitimate application in entertainment, education, & accessibility, its misuse presents significant legal & societal challenges. The use cases of A.I. have now turned into its demerits like the Deepfakes which have been used to facilitate financial fraud, spread misinformation, damage individual reputation through defamation, and create non-consensual sexually explicit content. It can now blur the distinction between reality & fabrications, threatening not only individual rights but also public trust in digital information.
Against this backdrop, an important legal question arises: Does India’s existing legal framework have the capacity to address deepfake-related crimes?
UNDERSTANDING DEEPFAKE CRIMES
The word “Deepfake” combines “deep learning” and “fake content”.[2] Deepfakes are media, especially videos or images where a person’s face or voice is swapped or altered using AI. The most common forms of deepfakes are generally fake videos, fake audio and fake images.
These deepfakes may generally look real and can be generated in the following steps:
- Collection of Data: AI models require source material. So, to target a person, creators collect hundreds to thousands of photos, videos, or audio recordings of the target person.
- Processing of Data: AI cleans and maps the given material by identifying facial landmarks (eyebrows, mouth, nose, eyes) and isolating vocal patterns from the given (collected) data.
- Model Training: There are two frameworks that creators use to generate the fakes: One is “Generative Adversarial Networks” (GANs) in which two AI models compete against each other. The “generator” creates fake content, while the “discriminator” evaluates it, and this cycle continues until it becomes indistinguishable from the original.
Another framework is “Autoencoders” in which AI encodes the target’s facial expressions, then decodes them to superimpose onto another image.[3]
- Generation and Post-Processing: Finally, system synthesizes the final image, video, or audio. By aligning all the features, creators smooth out the transitions for realism.
A recent example of a deepfake case is the Preity Zinta case, in which fake AI chatbots misused her identity. After this, she approached Bombay High Court, which granted interim protection and ordered removal of the fake content.
Another example is Anil Kapoor v Simply Life India, in which the Delhi High Court awarded Anil Kapoor with the protection against unauthorized usage of his image created by artificial intelligence. The use of such technology is considered a violation of personality rights.[4]
TYPES AND IMPACT OF DEEPFAKE CRIMES
The major type of Deep fake crimes is identity theft add financial fraud, where scammers impersonate company executives or family members using video clones or fake voice notes to steal money. Another form of deep fake crime is cyber bullying and harassment, in which creators generate non-consensual Altered images or scandalous fake videos to ruin the reputation of their target person.
Deepfakes Are also used in disinformation campaigns to manipulate public opinions and to spread fake news or making politicians appear to say things they never did.
Impacts of these crimes is severe, as it causes financial loss, damage of reputation, and emotional trauma. Because of these crimes, it becomes difficult to separate truth from lies in our daily lives as it erodes public trust in digital media.[5]
EXISTING LEGAL FRAMEWORK IN INDIA
India doesn’t have a specific deep fake law yet, but there are existing laws that can be applied depending on the offence to tackle the menace.
- Information Technology Act, 2000: The Information Technology Act, 2000 has various provisions that can be used to prosecute deepfake offences. For instance, it penalizes non-consensual capture or transmission of any intimate image, thus, applying to deepfake sexually explicit images. Moreover, the act criminalizes publication or transmission of any obscene material in electronic form. Cheating by way of impersonation through online platforms can also be prosecuted under this act, for example, deepfake financial frauds and identity theft.[6]
- Bharatiya Nyaya Sanhita, 2023 (BNS): The Bharatiya Nyaya Sanhita, 2023 can also be invoked where deepfakes are used to commit offences such as defamation, insulting the modesty of a woman, forgery of electronic records, or voyeurism involving manipulated images or videos. These provisions provide remedies against the misuse of deepfake technology, although the Act does not specifically recognize deepfakes as a separate offence.[7]
- Digital Personal Data Protection Act, 2023: The Digital Personal Data Protection Act, 2023 works alongside the Information Technology Act, 2000 and the Bharatiya Nyaya Sanhita, 2023 (BNS) to address deepfake-related offences. It protects Individuals personal data, including their photographs, voice recordings, videos and biometric information. In deep fake cases, using a person’s image or voice without their consent may violate the act. Although the act does not specifically criminalise deepfakes, it protects against the unauthorised processing of personal data and allows penalties for such misuse.[8]
CHALLENGES IN REGULATING DEEPFAKE:
The three major challenges in regulating deepfakes are – Lack of specific legislation as there is no law directly defining or punishing deepfakes. Secondly, anonymity of creators, as a default can be maintained and uploaded anonymously from anywhere. The third challenge is the jurisdiction issue, as servers can be in one country, creators in another and the victim in a third.[9] There are other challenges like the rapid advancement of AI technology, due to which there is increased use of AI in many fields which increases deep-fakes and crimes. Due to this much advancement in AI sometimes it becomes difficult to detect deepfakes.
SUGGESTIONS AND THE WAY FORWARD
To address the issue of deepfakes, legislation must be amended in several key areas. First of all, parliaments should adopt laws that impose severe penalties on the creation and distribution of non-consensual deepfake content. In addition, lawmakers must obligate developers to introduce technologies that allow detecting artificially generated computer content at the level of cryptographic provenance.[10]
Moreover, there is a need to equip law enforcement agencies with advanced technical equipment and provide investigators with special training to track down anonymous perpetrators. Another promising direction is compelling technology companies to create systems for quickly removing illegal content and introducing user-friendly reporting mechanisms. Finally, governments should invest in educating the public about digital security and training citizens to detect deepfakes. By combining swift legal responses, cutting-edge technologies, and digital literacy, society can ensure personal safety in the face of computer-generated fraud and protect the future of technologically driven innovation.
CONCLUSION
From being a novelty, deepfakes have evolved exponentially, becoming a serious threat to privacy, finances, and even democracy. Cases such as Preity Zinta’s highlight the potential of artificial intelligence (AI) to cause damage far beyond reputational harm. India does not have a specific deepfake law, but myriad statutes such as the Information Technology Act, 2000, the Bharatiya Nyaya Sanhita, 2023, and the Digital Personal Data Protection Act, 2023 can be leveraged to charge offenders with impersonation and defamation.[11]
Yet, the anonymity offered to deepfake creators, the jurisdictional challenges in prosecuting foreign offenders, and the rapid advancement of AI technology present formidable challenges to effective regulation. To mitigate these risks, it is essential to enact specific deepfake laws that impose stricter penalties, mandate cryptographic watermarking, adopt advanced tools for detecting synthetic media, and hold platforms accountable. Ultimately, the need of the hour is a balanced approach to deepfake regulation that promotes innovation while protecting individuals and institutions from exploitation.
Author(s) Name: Tanyia Bhagat (The Law School, University of Jammu)
References:
[1] UNESCO, Guidance for the Governance of Digital Platforms (UNESCO 2023).
[2] Robert Chesney and Danielle Citron, ‘Deep Fakes: A Looming Challenge for Privacy, Democracy, and National Security’ (2019) 107 California Law Review 1753.
[3] Ian J Good fellow and others, ‘Generative Adversarial Nets’ in Advances in Neural Information Processing Systems (MIT Press 2014).
[4] Anil Kapoor v Simply Life India and Others 2023 SCC OnLine Del 6914.
[5] World Economic Forum, The Global Risks Report 2024 (WEF 2024).
[6] Information Technology Act 2000, ss 66C, 66D, 67 and 67A.
[7] Bharatiya Nyaya Sanhita 2023, ss 336, 356, 294 and 351.
[8] Digital Personal Data Protection Act 2023, ss 2(t), 4, 5 and 8.
[9] Ministry of Electronics and Information Technology, Advisory to Intermediaries on AI and Deepfakes (15 March 2024).
[10] Organisation for Economic Co-operation and Development (OECD), OECD Framework for Classification of AI Systems (2022).
[11] Law Commission of India, Report No 282: Harnessing AI for Good Governance (2024).