S N Poojitha
Artificial Intelligence (AI) has moved from the realm of science fiction into something very engrossed in everyday fabric of life. Whether it is chatbots drafting emails, algorithms recommending news, coding apps or deep-fake videos simulating real people; AI now touches everything around us. With this power comes the question of responsibility: how should we regulate a technology that can create and distort reality? In India, the stakes are especially high—nearly a billion internet users (with cheapest data rates in the world), a vibrant digital ecosystem, young impressionable population and diverse socio-cultural fault lines. The government’s recent instructions on watermarking and labelling AI content mark one of the most significant ethical and regulatory interventions in the country’s history and has been a topic of hot discussion in tech circles.
The Ministry of Electronics and Information Technology (MeitY) has proposed draft amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. These amendments aim to bring what the ministry calls “synthetically generated information”—content created or modified by AI into a publicly transparent and accountable legal framework. Under these draft rules published in October 2025, AI-generated media must be clearly labelled as such, with visible identifiers or watermarks. For example, if a video is AI generated then the video must have a label indicating that is made using AI. Platforms with more than 5 million users (called Significant Social Media Intermediaries, SSMIs) are required to verify such user declarations, deploy technical measures to detect synthetic content, and ensure viewer is aware of it. This is very much akin to cigarettes boxes having marking on them indicating that they are harmful to health with only difference that in case of AI videos/images the viewer is given clarity that it is AI generated.
Primary reason for India to act now is the rapid rise of generative AI and its potential misuse being flagged by many citizens. AI-produced images and videos—commonly called deepfakes—can impersonate individuals, skew public opinion, facilitate fraud, and inflame communal tensions. India’s varied linguistic, cultural and political landscape makes it especially vulnerable. AI has been used to morph images, create content to misguide, blackmail or even malign many personalities.
The ethical issues go beyond mere labelling. At the heart lies a fundamental question: when does synthetic content become deceptive rather than creative. A comedian’s AI-generated spoof may be harmless, but a doctored communally sensitive video during elections could lead to unrest. The new rules also intersect with freedom of expression: how much should platforms and governments control what users create or share? By mandating user declarations and platform verification, the government hopes to strike a balance: protect citizens without stifling innovation.
Watermarking or labelling is the first layer of regulation, but it also brings practical challenges. Embedding visible markers is technically tricky and even more tougher is to ensure that such markers are not edited out again by mischief mongers.
From a regulatory ethics perspective, India is charting new territory. The draft measure that any watermark should cover at least 10 % of the screen or initial duration is reportedly among the first explicit “quantifiable visibility standards” globally.
The draft is still under stakeholder consultation and not yet final. It currently focuses on content that is “reasonably appears to be authentic or true” but has been created or modified by AI. (Some observers note that text alone may not be covered immediately, and the rules may avoid prescribing strict penal consequences in the first phase: an approach meant to encourage innovation
In practical terms, this means businesses and creators must prepare. AI-tool developers may need to embed invisible or visible watermarks. Platforms must implement systems for user declarations, metadata tracking and synthetic-content detection. For Indian creators, marketers and media houses, the new regime means that content generation and sharing will come with compliance burdens. For technology firms, the costs of automated detection and labelling will rise. For citizens, the promise is clearer disclosure: when you see a video or hear a recording, you’ll know if it’s genuine or synthetic.
On the plus side, these regulatory efforts promise greater trust in digital content, a reduced risk of manipulation during elections and better protection of citizen reputations. On the flip side, over-regulation could stifle experimentation, delay creativity and increase costs for small players. There’s also the risk of “label fatigue” where disclaimers lose meaning because they become too common or technical. The ethical frontier will lie in enforcement: if platforms fail or labels become a tick-box exercise, the value of the system falls.
Looking ahead, India is one piece of a global jigsaw. The EU’s “AI Act,” the United States’ executive orders and China’s own rules all show that AI regulation is entering a phase of regulatory competition. India’s unique challenge is to build rules that suit its vast diversity, linguistic multiplicity and digital scale: while preserving its startup ecosystem and innovation drive.
In conclusion, India stands at a crossroads in AI regulation and ethics. The flood of generative AI, deepfakes and synthetic media demands action: and the government has responded with a robust draft framework of watermarking, labelling and platform accountability. But technology evolves rapidly, and regulation must adapt in tandem. The ultimate measure of success will be whether the labels really work, platforms comply and citizens regain confidence in what they watch and hear. For India’s digital future to be both vibrant and safe, ethics must not be an afterthought.
***
