A Digital 'Jhakaas': What Denmark's Deepfake Law Means For India

Introduction

Imagine a world where your favourite actor's "Jhakaas" catchphrase is convincingly mimicked in a video, not by a skilled impressionist, but by a soulless algorithm. And now visualize that in the hands of someone who can generate the damaging, non-consensual video of you using the same technology, called deepfakes. It is not a science fiction, but our new reality. Deepfake is a worldwide crisis with claims that the number of deepfake videos has risen by more than 550 percent since 2019. Whether its shimmering bling of Bollywood or the peaceful European towns, this technology promises to chip away at the belief of the people, tamper with the political process and cause more severe individual damages.

This is an issue that India has been struggling with in the digital age, frequently via a chain of landmark cases that have tried to shoehorn new technological offenses into pre-existing laws. However, a recent bill proposed in Denmark suggests a radically different, affirmative plan of action. It also seeks to provide all citizens with a right of ownership over their faces, voices and bodies with a copyright-like right. This radical, inclusive solution is an innovative reaction to the overall threat, which reinvents this concept of identity in the digital world.

This blog post will examine India legal status of deepfakes, analyse the revolutionary Danish model, and explain how a copyright-like concept regarding personal identities as a counterpoint to other models worldwide such as the European Union AI Act could become the breakthrough to India in the virtual present. The main question that we shall address is whether or not India will maintain the hue of reactive and celebrity-focused nature, or shift towards more democratic and open umbrella to all of its citizens.

The Indian Reality

Indian stance on deepfakes has long been termed as a legal Wild West. The general lack of a dedicated law against synthetic media has left both the police and the victims to cobble together what little coverage non-specific laws towards that end provide, none of which was made in an AI generation context. This is a complicated and rapidly changing legal environment and a number of major-statured laws are invoked:

• The Information Technology (IT) Act, 2000: This legislation is the backbone of the Indian cyber law and many of its provisions have been strained to take into consideration the harms of deep falsities. When deep fakes are employed in order to conduct fraudulent schemes, Section 66D about cheating by impersonation ends up being used. Section 66E on the privacy offense would apply in the case of deepfake being generated based on the personal images or data of people without their approval. Moreover, Section 67 and 67A that addresses publishing of obscene or sexually explicit material is vital in cases of non-consensual deepfake pornography, as it creates severe punishment to the culprit.
• The Indian Penal Code (IPC) and the Bharatiya Nyaya Sanhita (BNS): Where deepfakes are created to damage the reputation of a person, sections 499 and 500 of IPC on criminal defamation have been employed. Even more specific is the newly introduced BNS, 2023, where its provisions target the curb of misinformation and disinformation as something that can lead to the mischief of the population as Section 353 suggests. Also, Section 77 of the BNS criminalizes the recording or distribution of an images of a woman doing a private act against her will and would also be a potent weapon against deepfake based voyeurism.
• The Digital Personal Data Protection Act, 2023 (DPDP Act):DPDP Act was the most important step of India to a complete data privacy system. It mandates Data Fiduciaries (along with other actors such as AI companies) to process any collected personal data according to the rule of law and unambiguous user consent; as well as introduce far-reaching security measures. Personal information used to make Deepfake content without permission will clearly be an invasion of privacy, and it could end up with a serious fine. One significant flaw of the Act however, as has been noted by legal experts is that consent requirements have been drafted by the Act to only include data that is not publicly available as it is most deepfakes which derive their source material as such. That creates a large legal vacuum and points out the necessity of a more targeted deepfake legislation.

Such dependency on a patchwork system of laws has been most evident in the judiciary. Indian courts have recently been utterly required to reckon with the new reality of AI generated content in a series of cases that have become landmark trials.

• Anil Kapoor v. Simply Life India and Ors.: A precedent-setting 2023 ruling involved the Delhi High Court furnishing a veteran actor an ex-parte interim omnibus injunction. The court specifically afforded him right to his personality whereby it sought to injure others who were using his name, voice, pictures and the iconic catchphrase of this nickname JhaKaas to make commercial purposes. It was a broad injunction that directs the transfer of the infringing domain names and also barring him to be portrayed as a motivational speaker.
• Arijit Singh v. Codible Ventures LLP: Bombay High Court in an historical judgment upheld personality right of the legendary singer. The court acknowledged that the personality rights of an individual are infringed when his/her voice is cloned by means of AI tools without the consent of such an individual. The decision was done in such a manner that the unique characteristics to his voice such as his speech style and mannerisms were guarded.
• Ankur Warikoo v. John Doe and Ors.: Popular personal finance teacher Ankur Warikoo won an injunction in the Delhi High Court against deepfakes that were being used to spread a stock market scam using his face. The case is also remarkable due to indicating the speed of the judicial response to AI-based financial crimes and the application of the so-called John Doe orders to track unknown criminals. Social media companies were also criticized by the court through their slow-paced action with regards to takedown requests.
• Rajat Sharma v. M/s Independent News Service Pvt Ltd: The Delhi High Court allowed a news anchor, Rajat Sharma to obtain a prohibitory injunction on grounds that his own deepfakes were used to sell ineffective medicinal products. The judicial case highlights the propensity of the courts to apply protection of personality rights to those who work with media, including journalists, or are social and other kinds of figures to a greater extent in the context when deepfakes in question create a threat to the lives and health of people.

Although these successes have set a strong precedent, it is very reactive in nature and only people with the good reputation and the funds to spend on costly and time-consuming court action can exploit it. To the ordinary citizen of India who probably lacks well established "personality rights," or the means to assert these rights, this policy does nothing to provide a proactive protection against actual harms.

The Danish Solution

Although these successes have set a strong precedent, it is very reactive in nature and only people with the good reputation and the funds to spend on costly and time-consuming court action can exploit it. To the ordinary citizen of India who probably lacks well established "personality rights," or the means to assert these rights, this policy does nothing to provide a proactive protection against actual harms.

This law had key provisions which are:

• Universal Protection: It brings universal usage of the protection of the digital identity by protecting all citizens, not only the high-profile citizens.
• Right to Takedown: It grants people the ability to request the instant removal of non-consensual deepfake material online.
• Platform Accountability: It makes technology businesses responsible, by issuing them with dire penalties in case they fail to take action on takedown requests.
• Explicit Exceptions: It also considers exceptions granted on reasonable grounds on parody and satire by balancing such right with the freedom of expression.

Such a strategy is radical as it generates a universal, enforceable right that precedes the actual creation of the deepfake. It puts the burden on the creator and the platform instead of the individual victim to give a strong holding on legal grounds to pursue takedown and compensation.

The European Union's Transparency Model

The European Union proposed the AI Act which is another influential governing model on deepfakes. In contrast to the rights-based approach of Denmark, EU AI Act highlights a tiered and risk-based approach and induced transparency. It classifies AI systems into four types of risk (unacceptable, high, limited and minimal) and assigns varying obligations to each.

The Act contains stringent requirements of transparency in the case of deepfakes and other generative AI. It requires that generative AI system providers have to mark the output of their systems in a machine-readable manner and detectable as an AI-generated or manipulated output. This makes an important difference. Beyond granting platforms and the populace the instruments to detect deepfakes, the EU approach would not create a certain kind of right akin to a copyright. This is in addition to promoting the adoption of technical watermarks and metadata, as well as other provenance aids; actions that are backed by efforts such as the Coalition for Content Provenance and Authenticity (C2PA) on the global scale.

What's Next for India?

Though India has already achieved admirable steps in addressing deepfakes by using judicial precedent, advisories to platforms, the new DPDP Act, the provision of a detailed solution in an accessible manner is still absent. Recent statements and reports by the government and Press Information Bureau (PIB) confirm that the ministry of Electronics and Information technology is fully aware of the threat and is taking actions to mitigate the same which may include a future Digital India Act.

The scenario in India which is reactive and celebrity centred and the proactive and more person-centred approach in Denmark throws up a very legitimate argument, whether we will keep living by an ad hoc of antique laws and a celebrity's driven precedence, or will we shift to a more democratic, universal approach to secure the digital self of every person. With the Digital India Act coming into shape, the decision will not merely be a definition of our legal system but crucially, the synthetical age itself. A midway option, where the rights-based system is combined with the requirement of transparency put forward by EU, might become the most efficient means of a safe and stable path, where all Indian citizens will be able to say "Jhakaas" instead of a privileged few.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.