The Indian intellectual property (IP) landscape is undergoing a quiet change. Various government institutions have been active on the IP front, issuing concept notes, circulars, and guidelines, making claims to protect 'innovation', create a 'level playing field', or 'align with global standards'.

Some readers may have heard the quip 'copyright protects expression, not ideas'. The principle being that copyright law protects authorial expression but not ideas, themes, or concepts that might underlie the said expression.

The content of these guidelines, notices, and circulars seeks to have the effect of increasing the scope of subject matter that is otherwise normally protected, or creating new instruments of exclusion to cordon and close off information goods as private property. This subject matter expansion is happening across various domains, from Graphical User Interfaces (GUIs) to pharmaceutical drugs to generative artificial intelligence (Gen AI) and across intellectual property (IP) laws. This expansion is unwarranted, unsupported by existing evidence, and methodologically flawed.

Some Theory

To clarify, IP laws on patents, copyright, regulatory data protection, and designs govern the production and distribution of information goods, which convey information or knowledge of something. Information goods can be anything, from a basic checklist to a chemical compound. To illustrate, take Risdipalm, a small-molecule drug that can treat spinal muscular atrophy (SMA), a rare disease condition. A patent covering the chemical molecule Risdipalm protects knowledge of the chemical molecule Risdipalm that has the ability to treat SMA. The patent does not protect the chemical molecule itself, but rather the knowledge of the said molecule. This understanding is necessary to appreciate that IP laws protect intangible information, not the tangible products or processes that embody the information. This logic also extends to designs, copyright, and regulatory data protection.

Similarly, not all information goodsare eligible for exclusionary protection. Some readers may have heard the quip—copyright protects expression, not ideas. The principle being that copyright law protects authorial expression but not ideas, themes, or concepts that might underlie the said expression. Thus, the Harry Potter novels as a work are protectable subject matter, while the more abstract concept, theme, or idea of a wizarding school located among normal non-magical human beings is available for everyone to use.

Both the abstract concept/themes along with expressive works like a novel are information goods, but only the novel is eligible for copyright protection, not the abstract concepts.The current IP regimes are an extension of the market regime that is found in capitalism into the arena of information. What kind of information goods come under the remit of IP laws in technical jargon is called subject matter eligibility.

Thus, an expansion in subject matter eligibility would mean that more information goods (which are non-rival and non-excludable in nature) are made eligible for private ownership, thus excluding others from their use. This is not optimal in itself because any state intervention through exclusionary rights to non-rival goods imposes access costs on consumers of said information and reduces competition.

Non-excludable here means that I cannot exclude someone from using a good. For example, if manufacturer A discovers a drug and starts selling it in the market, a competitor can always purchase the drug and reverse engineer it to identify the pharmaceutical compound and can start manufacturing it himself. Manufacturer A cannot stop his rivals from manufacturing the drug.

What happens is that granting exclusionary rights over information can lead to situations where individuals who are in need of products embodying protected information cannot access them because they are effectively priced out of the market.

This is where a patent regime comes in. It makes an otherwise non-excludable good into an excludable good by state intervention. If manufacturer A had patented the drug, then he could approach the court and seek an injunction and use the strong arm of the state to prevent or exclude others from manufacturing or utilising the knowledge of said drug.

A good is non-rival if the use value of the good does not diminish when the number of users increases. In other words, my reading of the Harry Potter novels does not take from you the ability to read the same Harry Potter novels, whereas my eating of an apple prevents you from eating the same apple. An apple is rivalrous, while novels are not.

Information goods, due to their non-rivalrous and non-excludable nature, are absent state intervention free to use. As Talha Syed and Oren Bracha point out, because information is non-rivalrous, uses do not conflict, extinguishing the primary justification for property: excluding others to protect one’s own use-value. Why, then, intervene at all? Across normative theories of IP, intervention is justified solely to secure delimited compensation—because generating information often requires high upfront costs while copying costs almost nothing, creators need a way to recoup their investments or receive a fair reward for their labour. Thus, compensation is necessary, but exclusion is not: the right to exclude is merely one institutional mechanism among many (such as state grants or prizes) to compensate creators.

Since information is non-rivalrous, the marginal cost of producing information is zero. Thus, an exclusionary regime imposes access costs and creates what the law and economics literature call 'deadweight loss', which prevents the full and effective use of the information good in question. What happens is that granting exclusionary rights over information can lead to situations where individuals who are in need of products embodying protected information cannot access them because they are effectively priced out of the market.

This reality is very stark in the context of access to medicines, where patented drugs can be priced at a very high rate, effectively pricing out most patients who need them. This is a good example of deadweight loss at play,where the deadweight loss is the actual loss of people’s lives due to inability to access the drug. This can play out with different forms of information goods, imposing varying harms.

No new evidence has been generated to support the expansion. Rather, it is vaguely supported by claims of protecting innovation, preserving the creator economy, or by the vaguer claim that India has to align with 'global IP standards'.

The entire debate then turns on a utilitarian calculus, an attempt to effectively calculate the optimal level of protection and to extract maximum bang for the buck.

What does all of this mean? IP rights impose access costs and reduced competition on consumers to enable information producers or owners to secure private value from the social value generated by the said information through a market-based exclusionary regime. There is no first-order normative principle supporting exclusionary rights over non-rival information; rather, as Syed and Bracha argue, exclusion serves only a second-order institutional purpose: instead of relying on centralised state planners to fund innovation, the state grants temporary exclusionary rights so producers can charge market prices, using decentralised consumer demand signals to channel what gets created. However, using market-based property as a tool within capitalism (a system of generalised market dependence) carries a rider. While IP is only meant to provide limited cost recovery, capitalist markets operate on the imperative of limitless profit maximisation and ceaseless expansion—driving rightsholders and lawmakers to convert every uncaptured social use of information into private exchange-value, inevitably triggering rights creep. Thus, any subject matter expansion regime must be seen in this light and must be supported by evidence that increased access costs and the decrease in competition that accompanies such expansion are outweighed by the benefits of 'innovation'.

This piece makes a methodological critique and argues that recently the government and its various organs, along with courts, have been expanding subject matter by granting new exclusionary protection over information goods and are contemplating or creating new mechanisms for exclusion without any additional justification, either empirical or otherwise. No new evidence has been generated to support the expansion. Rather, it is vaguely supported by claims of protecting innovation, preserving the creator economy, or by the vaguer claim that India has to align with 'global IP standards'.

This development creates a difficult dilemma. Decisions by both the courts and the current government are steadily expanding the scope of IP regimes absent justification, a consequence of the institutional structure of IP rights I discussed above. This trend triggers rights creep—the gradual expansion of legal protections for information goods driven by economic, political pressures, rather than sound policy justifications or strong legal and conceptual moorings—which ultimately inflates costs for consumers of information goods without delivering any measurable boost to innovation.

Looking back at legal history reveals that the traditional status quo was vastly different.

Change in Status Quo?

This was not the case earlier. India was only begrudgingly dragged into the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) in 1995. Even when it was dragged into TRIPS to introduce a broad definition of what constitutes an 'invention', it had also taken advantage of the TRIPS flexibilities to put limits on the scope of subject matter, in the context of introducing a detailed Section 3, including Section 3(d) in the Patent Amendments Act 2005.

Similarly, it was not long ago, as pointed out by Prashant Reddy T and Sumathi Chandrashekaran, that the Patents Bill introduced in 1965 and 1967 proposed a “deliberately weakened patent law” with reduced term limits and license-of-right regimes. The final 1970 Act did away with product patenting for food, drugs, and medicines.

It would be apt to borrow Basheer’s words that the patent script is a bit schizophrenic now, with various pushes and pulls acting on the decision-making.

In a lighter vein, Shamnad Basheer had in 2017 narrated the story of a former Joint Controller who stayed up all night to find ways to reject a patent! Similarly, in a paper titled Policy Style Reasoning, written in 2005, Basheer again discusses how the ghost of Ayyangar still haunts the Patent Office. This influence subtly shapes controllers’ decisions against granting patents and promotes a restrictive interpretation of the Patents Act, including on subject matter eligibility. In this way, it provides controllers with a certain ideological mooring that skews their decisions towards a restrictive reading of the act.

The ghost in question is Justice N. Rajagopala Ayyangar, whose report was the basis on which the Patents Act 1970 was passed. The Ayyangar Committee report, submitted in 1959, was a well-researched policy document whose major recommendations included doing away with product patenting for food, drugs, and medicines, and advocated a cautious and sceptical approach to patent law.

But things have drastically changed. Prashant Reddy, roughly 18 years after Basheer’s 2005 paper, narrates a different story about the Indian Patent Office. It would be apt to borrow Basheer’s words that the patent script is a bit schizophrenic now, with various pushes and pulls acting on the decision-making. This can be seen in the behind-the-scenes developments to do with the Patent Amendment Rules 2024.

Expand, Expand, and Expand

To list a few of the recent developments, on 23 January 2026, the Department for Promotion of Industry and Internal Trade (DPIIT) issued a concept note proposing amendments to the Designs Act, which broadly advocated for protecting GUIs, icons, and virtual goods under the Act. As I have pointed out elsewhere, GUIs are protected under three distinct regimes though they should not be protected under any of them. It has also been pointed out that GUIs are the ideal poster boy of IP expansionism in the absence of any legal conceptual moorings to constrain such expansion.

Similarly, on 8 December 2025, a Working Paper on Generative AI and Copyright was released. It advocated a statutory licensing model for AI model training, potentially again expanding the scope of subject matter and including uses and information goods that are otherwise not protected by copyright law. The Central Drugs Standard Control Organisation (CDSCO), in its notice dated 8 October 2025, sought to create a level playing field between first mover and follow-on manufacturers on clinical trial data, which establishes the safety and efficacy of a drug.

A wave of recent developments has expanded the scope of IP protection to include new subject matter, often without any clear justification.

The notice seems to indicate it was in favour of some form of data exclusivity regime that would protect the clinical trial data generated by first movers and prevent follow-on manufacturers from relying on the said data for a few years. Even though the last development does not constitute subject matter expansion, it is nonetheless seeking to extend some form of exclusionary protection over clinical trial data, an information good that is otherwise only protected under a data secrecy regime and not an exclusionary regime.

This is not all. Courts have not been sitting idle and have been pretty active in expanding subject matter through their own decisions. To illustrate, the recent decisions emerging from the Delhi and Madras High Courts with respect to Section 3(k) of the Patents Act 1970 have reduced the section from a substantive objection to a mere formality.

Section 3(k) bars patenting over software, business, and mathematical methods. Unfortunately, the section says “computer programme per se.” There has been a huge debate about what this “per se” means. The courts in recent times have come out with a slew of decisions interpreting this section, and each of its decisions has reduced the scope of Section 3(k), reducing its ability to act as an independent standalone objection. This has paved the way for software patenting.

The Calcutta High Court in NEC Corp v. Controller did something similar, but this time under the Designs Act, expanding the scope of design law to include GUIs as eligible subject matter to be protected. One could list many similar examples, but these are enough to show that the landscape of Indian IP has fundamentally changed. A wave of recent developments has expanded the scope of IP protection to include new subject matter, often without any clear justification.

Whither Data?

The consensus in IP law is that the system must be carefully calibrated to deliver the maximum social and economic benefit. To achieve this, the IP regime must strike a balance between rewarding innovation and ensuring public access.

But as has been pointed out by Mark Lemley and William Fisher, evidence for the justification of the existence of the IP system is barely forthcoming. In a sense, one is caught in the middle. It is unclear what the specific benefits and harms of the existing IP system are, and whether one outweighs the other. In other words, no one is able to say whether the benefits of the existing IP systems outweigh the costs that accompany them, or vice versa.

These publications merely offer explanatory data or highlight the potential negative effects of the current ecosystem. They fail to address fundamental questions, such as whether India should expand IP protection to encompass new digital information goods.

It would then make ample sense to maintain the status quo and neither expand nor contract the scope of IP rights as they currently exist until coherent reasons are provided. But, unfortunately, as can be seen, that is not the case.

In India, the issue is more serious. The Indian IP regime is woefully data-poor—very little empirical work examines how India’s IP system actually functions. To illustrate this point, take a recent Kerala High Court decision involving Ribociclib (a breast cancer drug). The court chose not to issue directions to the union government to exercise government use provisions under the Patents Act for the simple reason that there was no adequate data on  “number of diagnosed patients; patients consuming medicines and those not consuming them [Ribociclib] due to high cost; and patients who have benefited from existing health schemes.”

This does not mean there has been no empirical work. Studies have been done on the impact of introduction of product patenting on generic competition, in addition to Sunil Mani’s article discussing patent filings based on data from the Intellectual Property Office (IPO). The IPO, in its annual reports, along with the World Intellectual Property Organization (WIPO), puts out some interesting statistics on the functioning of the IP regime in India. However, these publications merely offer explanatory data or highlight the potential negative effects of the current ecosystem. They fail to address fundamental questions, such as whether India should expand IP protection to encompass new digital information goods.

As noted elsewhere (see here, here, and here), none of the developments mentioned in the previous section seems to provide a coherent rationale for introducing more exclusivities over information through IP laws. Consider Section 3(k) and software patents. Ferid Allani v. Union of India heralded a sea change in Section 3(k) jurisprudence. In this decision, the Delhi High Court observed that withholding patents for software in today’s digital world would be a retrograde step.

However, this approach raises critical concerns. First, it is debatable whether courts should engage in such policy-driven reasoning. Second, the mere ubiquity of a technology does not automatically justify cordoning it off under a restrictive patent regime.

Similarly, take the concept note that was issued by the DPIIT in the context of GUIs, icons, and virtual designs. That 92% of the countries in the world protect GUIs and virtual designs in some way was seen as justification to introduce similar protection in India. Unfortunately, there was no analysis of its possible effects or a discussion of any empirical evidence that warrants such an expansion.

The central question remains—why are India’s executive and judicial bodies pushing to expand private monopolies over knowledge or information without empirical justification?

Finally, take the Working Paper on Generative AI and Copyright Law. This paper proposes a statutory licencing regime without first resolving the fundamental question of whether training generative models on expressive works constitutes copyright infringement. While the regime is ostensibly designed to protect the Indian creator economy and maintain production incentives, the paper offers no empirical evidence to support its claims. It completely fails to explain how such a system will achieve its goals, or why a licencing regime is the correct policy tool to address the true underlying issue—the displacement of human creative labour by machines.

There have already been arguments made that copyright is not even the right tool for addressing the impact of GenAI on creative labour (see more on this here and here). None of the concept notes, notices, working papers, or court decisions have presented any empirical information or a coherent rationale, or both, to justify the subject matter expansion they advocate. Instead, they merely contain vague claims to either protect “innovation”, create a “level playing field”, “protect the creator economy”, or “align with global standards”.

The central question remains—why are India’s executive and judicial bodies pushing to expand private monopolies over knowledge or information without empirical justification? Across pharmaceuticals, software, design, and GenAI, policymakers are actively closing off public information goods based on wafer-thin assumptions—vaguely invoking “innovation” or “global standards” as a substitute for evidence.

When the state builds an exclusionary runway for information without data, the costs are paid by everyone. As the late legal scholar Basheer reminded us years ago, intellectual property policy cannot be built on mere faith; instead, it must be driven by empirical facts.

Acknowledgments: Thanks to Swaraj Paul Barooah and Akshat Agrawal for discussions on this topic and to Kriti Gwal for her assistance. The views expressed here are the author's own.

Bharathwaj Ramakrishnan is a patent lawyer.