Author: Cheshta Vashishtha
College: Narsee Monjee Institute of Management Studies, Chandigarh
Abstract
The saying “seeing is believing” has been a long-held belief. This old saying is quickly taking its meaning in the age of Artificial Intelligence. A computer-generated image can now show an individual saying words that were never spoken by him, a film can show an event that never really happened and a sound recording can perfectly mimic a person’s speech without his or her permission. Deepfakes are a type of AI manipulation that’s changing the face of deception in the digital world.
The incidents of deepfakes have been a growing problem in India, with cases involving identity theft, financial fraud, creation of explicit images without consent, and fake news. The country has welcomed AI as a tool for innovation and economic growth, but not so fast their laws. While the existing laws provide piecemeal remedies, there is still no specific legislation which provides for the regulation of deepfakes solely in India.
The Rise of a New Digital Threat
Imagine you get a video call from your parents and they have an emergency such as being stuck in another city and need money. The face is recognisable, the voice is the same and the fear is real. In minutes, you send the cash, only to find out at the end that the call was completely AI-generated.
This isn’t fictional anymore.
Deepfakes are synthetic media made using deep learning algorithms that can replicate facial movements, voice, words and gestures of a person with high accuracy. Now, AI can create content that is so convincing, even separating fact from fiction has become a challenge.
Technology is not itself a bad thing. It has a number of valid uses in filmmaking, education, accessibility, healthcare, and language translation. Its misuse has increased as fast, however. The impact of deepfakes has been felt in the realm of cyberbullying, identity theft, financial fraud, corporate fraud and electoral disinformation.
India’s fears heightened when a fake video of actress Rashmika Mandanna came out allegedly created by artificial intelligence. The initial video was not created by this person, but the AI technology has been used to superimpose the actress’ face onto it, causing a lot of public outrage and leading the Government of India to issue guidelines urging Digital platforms to increase protection against synthetic media.
The incident was a cautionary tale that deepfakes are not just a celebrity problem. Everyone who has an online presence is at risk.
The Legal Framework in India: A Jigsaw with missing pieces
At the moment, there is no specific law in India that addresses deepfakes exclusively. Rather, victims are forced to depend on a mix of constitutional rights and laws covering only a limited range of outcomes of AI-based deception.
Article 21 of the constitution provides for the right to life and personal liberty, and the Supreme Court has been interpreting that right to include the right to privacy, dignity and reputation. At the same time Article 19(1)(a) provides freedom of speech and expression and Article 19(2) allows for reasonable restrictions for the purposes of “public order, decency, morality, defamation, and national security”.
There are some remedies available under the Information Technology Act, 2000 in terms of Section 66C (identity theft), Section 66D (cheating by personation with the use of computer resources), Section 66E (violation of privacy) and Sections 67 and 67A (publication or transmissions of obscene and sexually explicit electronic material). But these laws have been in place way before the advent of complex Generative AI and do not explicitly control synthetic media.
The Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, introduce the concept of ‘due diligence’ for intermediaries, mandating them to take action on complaints of unlawful content. In a bid to curtail the surge in such incidents, the Government has also issued notifications to the intermediaries, urging them to take immediate action on content created using AI that is harmful to society.
The principle of consent in relation to processing of personal data has been reinforced in the Digital Personal Data Protection Act, 2023. However, it doesn’t contain explicit provisions regarding the production of AI-generated copies of a person’s voice, face or likeness.
Likewise, the Bharatiya Nyaya Sanhita, 2023 can be used when deepfakes are used as a means of committing offences like forgery, cheating, defamation, obscenity, criminal intimidation or personation. However, the provisions are only meant to deal with harm after it has been caused, not to prevent its creation or spread.
Hence, legal action is still reactive in India instead of preventative.
Constitutional Principles and Judicial Guidance
Although India has not yet witnessed a landmark Supreme Court judgment dealing exclusively with deepfakes, constitutional jurisprudence already provides valuable guidance.
The Right to Privacy was accorded the status of a fundamental right by a nine-judge Bench in Justice K.S. Puttaswamy (Retd.) v. Union of India (2017). This judgment sets a constitutional precedent for safeguarding people from misuse of their images, voices, or identities by AI, as such deepfakes are often created when the person’s image, voice, or identity is used without their permission.
Finally, it is important to mention the case of Shreya Singhal v Union of India (2015), where the Supreme Court ruled that Section 66A of the Information Technology Act was unconstitutional but also declared that any restriction on freedom of expression must meet constitutional norms of reasonableness and proportionality. The ruling is a reminder to lawmakers that future legislation to regulate deep fakes should not go too far and impede legitimate expression.
In R. Rajagopal v. State of Tamil Nadu (1994), the Court had held that an individual has the right to take steps to protect his/her privacy and honour from being misrepresented. Both these interests are directly threatened by deepfakes because they are presented as reality instead of being. Deepfakes pose a direct threat to both of these interests as they are presented as reality instead of being.
Together, these rulings prove that, at this point, Indian constitutional law ensures the safeguarding of values most at risk from deepfakes. Where there is a lack is in the legislation that is specific to the deception caused by AI.
Why India Needs a Dedicated Deepfake Law?
Deepfakes are so sophisticated that they require a legal answer beyond the realm of criminal law.
Any comprehensive law should start off with a clear definition of what constitutes a “deepfake” and clarify the difference between malicious synthetic media and artistic, educational, journalistic or satirical media. This would eliminate legal uncertainty and preserve freedom of speech.
Second, there must be more accountability by technology firms and social media intermediaries. Increased accountability for the creators of AI tools, labelling or watermarking requirements for AI-generated content and time-limited takedown policies can effectively curb the spread of misleading media, while not stifling innovation.
The victims must also be provided with quick grievance redressal systems, monetary damages for any damage done to their reputation, emotional damages and severe punishment for financial fraud, sexual exploitation, identity theft or electoral manipulation.
International developments can also be a source of lessons for India. The EU’s AI Act was already introduced in 2024 with a phased approach and is based on a risk-based regulatory framework, it also has transparency requirements for specific types of AI content, including deepfakes. The EU frameworkis an example of how innovation and accountability can be intermingled within a single regulatory model, although the social, constitutional and technological context of India is very different.
The Way Forward
The need to ensure digital literacy is a national priority. Citizens need to think critically about content on the Internet rather than taking everything they see online at face value, even if it looks convincing in video and/or audio format. There needs to be a collaborative effort between educational institutions, technology companies, and the government to raise awareness about the dangers of misinformation created by AI.
In tandem, AI can be the solution. New tools for content detection, digital watermarking, content authentication systems, and forensic AI have the potential to help investigators and digital platforms identify manipulated content before it gets to millions of users.
Never seeking to discourage innovation. AI has the power to revolutionize sectors such as healthcare, governance, education, and scientific research. Technological progress, therefore, should not be accompanied by a restriction of constitutional rights and at the same time, regulation must enable responsible innovation.
Conclusion
Deepfakes have evolved from a tactic that can be used for manipulation in limited moments to a high-tech tool that can cause a threat to privacy, democracy and public trust. With an ever-digital society, it’s more important than ever to tell the truth from the fiction, a legal and democratic imperative.
It is an important moment in India’s history. Its legal protections need to keep pace with the rise of its AI and digital innovation capabilities as a global leader. An all-encompassing deepfake legislation, grounded in constitutional principles, technological know-how, platform accountability, and robust enforcement, can span the current regulatory void without hampering innovation.
The biggest problem now is not to fabricate believable lies, it’s to assure the public that the truth is believable. When seeing is no longer believing, Indian Law needs to make justice clearly real.
