India Is Still Fighting to Define AI Copyright. Google Just Filed a Patent That Makes That Fight Urgent.

While India's courts and Parliament are still debating whether AI-generated content can attract copyright protection at all, Google has quietly moved on.

On 27 January 2026, the United States Patent and Trademark Office granted Google LLC patent US12536233B1. It describes a system where Google's AI evaluates your website, decides it is not good enough and replaces it serving users a Google-generated version of your page built from your own content, your brand identity and the user's personal search history.

India has no specific law to address this. No AI regulation. No amended copyright framework that accounts for generative AI. No data protection authority with full operational powers. And a Competition Commission that is still developing its understanding of algorithmic self-preferencing.

We are not unprepared because this technology is new. We are unprepared because we spent the last five years debating whether AI can own copyright instead of asking the harder question: what happens when AI uses your copyright to replace you?

What the Patent Does: In Plain Terms

Google's system works like this.

When a user searches for a product or brand, Google generates a search result page. It then scores the brand's landing page against a proprietary formula evaluating conversion rates, bounce rates, design quality and content quality. If the score falls below a threshold that Google sets unilaterally, the user does not reach the brand's page. They reach a Google-generated alternative built to look and function like the brand's page, populated with the brand's products, images and information, personalised using the user's previous search history.

That generated page can be shared with competing organisations. It can be stored and reused for future searches.

The brand receives no notice. There is no opt-out described. There is no appeal mechanism. Google decides your website is inadequate. Google replaces it. Google uses your content, your brand and your customer's data to do so. And Google can give what it built to your competitor.

This is not paraphrased or interpreted. This is what the patent document states.

India's Copyright Law Was Not Written for This

The Copyright Act, 1957 is the foundational statute. It was last significantly amended in 2012 before generative AI existed as a commercial reality, before large language models could produce original-seeming content at scale and before any platform had the technical capability to ingest a website and produce a functional substitute for it.

Under Section 13 of the Act, copyright subsists in original literary, dramatic, musical and artistic works. A landing page its copy, its visual design, its information architecture qualifies across multiple categories. The copy is a literary work. The design elements are artistic works. The structure and layout, in sufficiently original form, attract protection.

The problem is authorship. Section 17 vests copyright in the author, and Section 2(d) defines the author of a computer-generated work as the person who causes the work to be generated. When Google's AI generates a page using your content as input, who caused that work? Google's engineers who built the system? The system itself? The user whose search query triggered it? Indian copyright law has no clear answer because it was not written to address a system where the input is someone else's protected work and the output is a commercial substitute for it.

The courts have begun to approach the edges of this problem, though not its centre. ANI Media Pvt Ltd v OpenAI Inc & Anr (CS(COMM) 1028/2024), currently before Justice Amit Bansal of the Delhi High Court, is India's first and only generative AI copyright case. It was filed in November 2024 by Asian News International alleging that OpenAI trained ChatGPT on ANI's copyrighted content without permission. The court has framed four questions for consideration: whether storage of copyrighted content for AI training constitutes infringement; whether generating AI responses from that content infringes copyright; whether Section 52 fair dealing applies; and whether Indian courts have jurisdiction given OpenAI's US-based servers. Two amici curiae Dr. Arul George Scaria of NLSIU Bengaluru and Advocate Adarsh Ramanujan have been appointed. Arguments were still ongoing as of early 2026. No substantive ruling has been delivered.

That case involves AI training on content. Google's patent involves AI generating a commercial substitute for a brand's website and serving it to users. The legal question is different and harder. The ANI case is a perimeter dispute. 

The Intermediary Liability Dimension

India's Information Technology Act, 2000 and the IT (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 created a framework for platform liability designed for a different kind of platform behaviour.

The safe harbour under Section 79 of the IT Act protects intermediaries from liability for third-party content when they act as neutral conduits hosting, transmitting or providing access to content without initiating the transmission or modifying the content. Google has historically relied on this protection across its services.

The conduct described in this patent is not neutral conduit behaviour. Google's system is not hosting your content or transmitting it. It is evaluating it, scoring it, replacing it with an AI-generated alternative and serving that alternative to users who believe they are reaching your website. That is active creation and substitution not passive transmission.

In  2025, the Ministry of Electronics and Information Technology stated in Parliament that web scraping by intermediaries for AI training purposes is regulated under the IT Act 2000, specifically Section 43, which penalises unauthorised access and extraction of data. If scraping for training is regulated, the generation and commercial deployment of content derived from that scraping is a conduct the IT Act's intermediary framework was not designed to immunise.

The doctrinal argument that the conduct described in this patent falls outside the Section 79 safe harbour is serious. Whether Indian courts would agree, and how quickly a test case could produce a definitive ruling, is a separate question. The timeline matters enormously here.

The Competition Commission and Algorithmic Gatekeeping

The Competition Commission of India has been developing its understanding of digital markets steadily, if more slowly than European regulators. In October 2022, the CCI passed a landmark order imposing a penalty of Rs 1,337.76 crore on Google for abusing its dominant position in the Android Mobile Device Ecosystem, finding violations of Section 4(2)(a)(i), 4(2)(b)(ii), 4(2)(c), 4(2)(d) and 4(2)(e) of the Competition Act, 2002. The NCLAT upheld this penalty in March 2023 in Google LLC v Competition Commission of India (Competition Appeal (AT) No. 1 of 2023). The CCI separately imposed Rs 936.44 crore for Play Store policy violations.

Those orders established that Google is dominant in the relevant Indian markets and that it will use that dominance to protect and expand its position across adjacent markets. The CCI's own words are relevant: the central question, it noted, is "to what extent should a platform operator set governance rules purportedly to 'protect' its ecosystem and should there be limits to this self-assumed role."

A system that evaluates third-party brand websites against proprietary criteria, substitutes Google's own AI-generated version when those criteria are not met, and routes user traffic to Google's interface rather than the brand's website is precisely the kind of self-assumed governance role the CCI has been asked to limit. The conduct maps directly onto Section 4  imposing unfair conditions, limiting market access, and leveraging dominance in one market to protect or enter adjacent markets.

The Competition (Amendment) Act, 2023 strengthened the CCI's framework for digital markets. But unlike the EU's Digital Markets Act, which designates specific gatekeepers and prohibits self-preferencing by statute, India's framework requires the CCI to develop the analysis from first principles in each case. That is possible. Under the Android precedent, the doctrinal foundation is available. The question is whether the CCI will develop it fast enough to matter.

Digital Personal Data Protection Act, 2023: Enacted But Not Operational

India passed the Digital Personal Data Protection Act  2023. In principle, it addresses several of the concerns this patent raises directly.

The DPDPA requires personal data to be processed only for the purpose for which consent was given, with data principals informed of how their data is being used. A user's search history is personal data. Processing it to generate a commercial landing page for a third-party brand and sharing that generated page with competing organisations would, if the Act were operational, raise serious questions about consent, purpose limitation and transparency.

But the DPDPA is not yet operational. As of the date of writing, the Data Protection Board of India has not been constituted and the rules under the Act have not been finalised. The consent framework, the enforcement mechanism and the penalties regime all exist on paper only.

This is the structural problem India faces with this patent. The law that might constrain it exists. The institution that would enforce it does not yet function.

The Stalled Legislative Response

The Parliament's 161st Report of July 2021 the Standing Committee on Commerce's review of India's Intellectual Property Rights regime recommended amending the Copyright Act, 1957 and the Patents Act, 1970 to incorporate emerging technologies of AI and AI-related inventions. That recommendation was made four and a half years ago. No legislative amendment has followed.

In 2024, the Ministry of Commerce and Industry stated in Parliament that the existing IPR regime is "sufficient" to protect AI-generated works and that no new legislation is required. The Ministry pointed to existing provisions as adequate. Google's patent is a direct test of whether that confidence is warranted.

The Digital India Act announced by MeitY in March 2023 as a comprehensive replacement for the IT Act, 2000 was intended to address platform regulation, intermediary liability and emerging technologies including AI. It has not been introduced in Parliament, no public draft has been released, and by December 2025, MeitY Secretary S. Krishnan confirmed publicly that the government is unlikely to introduce a new digital law, preferring to rely on existing frameworks including the DPDPA and IP legislation.

What this means in practice is that the legislative architecture India would need to respond to a product built on this patent if one is launched will have to be assembled from a 1957 copyright statute, a 2000 technology act, a 2002 competition law and a 2023 data protection act that is not yet enforced. Every one of those instruments is capable of being applied. None of them was written with this scenario in mind.

What Indian Brands and Publishers Are Actually Facing

Indian e-commerce brands, regional publishers, news organisations and small businesses that depend on search traffic are in a structurally vulnerable position.

Their landing pages, built and maintained at significant cost, would be evaluated against Google's quality criteria. Many would fail not because their content is poor, but because their conversion rates, page speeds and design quality do not meet thresholds set by a foreign platform using metrics designed for a different market context. Indian regional language publishers, vernacular news sites and SME product pages are particularly exposed because they are less likely to have the resources to continuously optimise against Google's evolving standards.

When those pages fail Google's threshold, their users would not reach their site. They would reach a Google-generated alternative built without consent, served without notice, potentially shared with competitors. The brand loses the user relationship. The publisher loses the page view. Google gains the commercial interface.

Under current Indian law, the tools to challenge this in real time are either underdeveloped, unoperationalised, or unavailable.

The Internet Architecture Argument

Beyond the legal frameworks, there is a structural argument about what kind of internet India wants.

India has over 900 million internet users. It has one of the fastest-growing digital economies in the world. Its policy ambitions articulated through the Digital India programme and the National Data Governance Framework envision an internet that serves Indian users, Indian businesses and Indian interests.

A patent like US12536233B1 represents the opposite architectural vision. It is a vision where a single platform foreign, dominant, algorithmically opaque becomes the arbiter of which websites are good enough to receive their own traffic. Where the commercial layer between a user and a brand is owned not by the brand but by the platform. Where the open web is progressively replaced by a platform-mediated web in which the platform generates the content, controls the interface and captures the commercial relationship.

That is not a neutral technical development. It is a political economy question. And India's response through law, regulation and policy will shape what the Indian internet looks like for the next decade.

What Needs to Happen

Finalise and operationalise the DPDPA immediately. A data protection law without a functioning Data Protection Board is a statement of intent, not a legal protection.

The CCI should develop and publish a framework specifically addressing AI-driven self-preferencing in search markets. The Android precedent established Google's dominance and the CCI's willingness to act. The analytical framework for the next category of conduct needs to be developed before a product launches, not in response to one.

The Copyright Act needs amendment to address the authorship gap for AI-generated commercial substitutes and the question of what constitutes infringement when a platform ingests protected content to generate a competing product. The 2021 Standing Committee recommendation has been sitting unimplemented for nearly five years.

Indian brands and publishers need to act in the legislative gap now. Document your creative process. Register copyright where registration strengthens your position. Review your terms of service with explicit reference to AI ingestion and content substitution. Build the factual record that litigation or regulatory action would require.

The Fight We Should Have Been Having

India has spent years arguing about whether an AI can be listed as an inventor or an author. It is the wrong fight.

The right fight is about what happens when AI is used by a dominant platform to displace human authors, human brands and human businesses from the commercial relationships they built. That fight is not about AI rights. It is about human rights the right of a brand to control how it presents itself to its customers, the right of a publisher to monetise the content it creates, the right of a user to know when the experience they are having was generated by a platform rather than the business they chose to visit.

Google's patent US12536233B1 is a document about power. About who gets to decide what the internet looks like, whose content reaches users, and who owns the commercial layer of the web.

India has the legal tradition, the institutional capacity and the market scale to answer that question on its own terms. The 2022 Android orders proved that. What it needs now is the urgency to act before the answer is made for it.