Every patent system in the world asks an applicant three questions: is it new, does it involve an inventive step, and can it be applied in industry. The Patents Act 1970 asks those questions too, but it asks a fourth one first, and the fourth question is the one that gives Indian patent law its character. Before novelty is examined and before inventiveness is argued, the Indian statute asks whether the thing presented is the kind of thing the law is prepared to call an invention at all. A whole list of subject matter, set out in Section 3 of the Act, is declared not to be inventions, however new it may be and however cleverly it may have been arrived at. A new form of a known medicine that adds nothing to its efficacy, a scheme for doing business, a method of treating the human body, a plant variety, a computer programme standing alone: each of these can be ingenious, each can be commercially valuable, and each fails at the threshold because Parliament decided in advance that no monopoly would be granted over it. India, in short, patents by exclusion. The grants are conventional. The refusals are the policy.

That design was not an accident of drafting, and it was not borrowed from any other country’s statute book. It was the product of a deliberate argument, conducted in the first two decades after independence, about what a patent system is for in a poor country: whether it exists to reward whoever invents, wherever they are, or whether it exists to build domestic industry and keep essential goods within the reach of the population. The Act of 1970 answered that question in a particular way, abolishing product patents for food, medicines and chemicals altogether and leaving only process patents in those fields, and even after the amendments of 2005 restored product patents across all fields of technology, the exclusion list survived and grew. To understand the statute a reader has to hold both halves of it in view at once: a grant machinery that is recognisably the same machinery found in London or Munich, and a set of refusals, oppositions and compulsory powers that exist because India never accepted that the grant machinery should run on its own logic without a public-interest brake.

The law the Patents Act 1970 replaced

The statute that governed patents in India for the first quarter century of independence was the Indian Patents and Designs Act, 1911, a colonial enactment modelled on British practice. It offered product patents and process patents alike, a term of sixteen years, and a framework designed for an economy in which the typical patentee was a foreign manufacturer and the typical Indian firm was a distributor. Its own ancestry ran back through a line of colonial statutes to the Act of 1856 that first offered exclusive privileges for inventions in India, and its administrative heart was the same Controller-centred office the later Act retained, examining, granting and registering within a system whose case law looked to England for authority. What independence changed was not the machinery but the question asked of it: a legislature answerable to Indian voters, facing Indian prices, was bound to ask whether machinery designed for an imperial market served a developing one, and the two decades of inquiry that followed were that question being worked through with evidence rather than slogans. The record of that regime, as the official inquiries of the 1950s documented it, was that the overwhelming majority of Indian patents were held abroad, that patented medicines were priced for the markets of the countries that produced them, and that the existence of a product patent on a drug in India commonly meant that the drug was imported at a monopoly price rather than made in India at all. A patent system of that shape transferred income out of the country and built no industrial capacity in return, which is why the case for redesigning it was made not by radicals but by committees of judges, civil servants and industrialists working through the ordinary machinery of the state.

Why did India refuse product patents for medicines for over three decades?

A committee chaired by Justice N. Rajagopala Ayyangar, appointed in 1957 and reporting in 1959, examined the 1911 Act and recommended that product patents for food, medicines and drugs be abolished and only process patents retained in those fields, because product patents in a developing economy had produced high prices and little domestic manufacture.

The Ayyangar Committee was not the first official inquiry to reach for reform, and its authority is strengthened by knowing that a predecessor had been more cautious. The Patents Enquiry Committee under Justice Bakshi Tek Chand, appointed in 1949 within two years of independence, had recommended retaining the patent system with safeguards against abuse, and a bill introduced in 1953 on the strength of its report was allowed to lapse. The Ayyangar inquiry, a decade later and with a further decade of the 1911 Act’s operation to inspect, went the further distance, and the difference between the two reports is itself instructive: the case for abolishing product patents in the sensitive fields hardened as the evidence accumulated that the foreign-held product patent in India characteristically served an import monopoly and nothing else. Even after Ayyangar, enactment took more than a decade: a bill of 1965 lapsed, a bill of 1967 went to a joint committee of Parliament whose deliberations shaped the final text, and the statute that emerged in 1970, substantially amended in passage, was the product of that long gestation rather than of any single government’s programme. Statutes with this kind of ancestry, argued over for twenty years before enactment, tend to carry their purposes on their face, and the 1970 Act does: its exclusions and its public interest powers are not encrustations on a neutral grant machine but the negotiated residue of the longest legislative argument independent India has conducted about any item of commercial law.

The Ayyangar report is the intellectual foundation of the 1970 Act, and its reasoning deserves to be stated precisely because later argument keeps returning to it. A process patent monopolises one route to a product; a product patent monopolises the product itself, by whatever route it is made. Where a country’s firms lack the capital and research base to originate new medicines but are perfectly capable of devising alternative processes for making known ones, a process-only regime lets domestic industry compete on ingenuity of manufacture while the product patent regime forecloses the field entirely to the first filer. Justice Ayyangar’s committee concluded that for food and medicines, articles whose price determines whether people eat adequately and whether the sick are treated, the foreclosure was indefensible: the patent system should reward the Indian chemist who found a cheaper way to make a drug, not arm the foreign originator with a right to keep the drug out of Indian manufacture for the life of the patent. The Patents Act 1970, enacted as Act 39 of 1970 and brought into force on 20 April 1972, gave effect to that recommendation through Section 5, which denied product patents for substances intended for use as food, medicine or drugs and for substances produced by chemical processes, while leaving process patents available in exactly those fields.

What followed is the part of the story the Act’s critics and defenders both treat as established. Relieved of product patents in pharmaceuticals, Indian firms built the process chemistry skills the committee had predicted, learning to reverse-engineer molecules and design around originator processes, and over three decades that capacity grew into a generic pharmaceutical industry that supplied inexpensive medicines across the developing world. The same three decades are described by the originator industry as a period in which research-based companies had no reason to launch products early in India, or to invest in Indian research directed at Indian diseases, because the fruits of any successful research programme could be copied lawfully the moment they appeared. Both descriptions are true, and the tension between them is the tension the statute was built to manage. The reversal of the process-only regime in 2005, when India’s obligations under the TRIPS agreement required product patents in all fields, is the subject of the amendment trail article in this cluster and is treated there in full; what matters for this pillar is that 2005 restored the product patent while leaving the 1970 Act’s other distinctive machinery, the exclusions, the two oppositions and the compulsory licence, intact and in several respects strengthened.

How the Act is put together

A reader opening the Patents Act finds a statute arranged in the sequence of a patent’s life. It opens with definitions, turns immediately to the exclusions that declare what is not an invention, and then follows an invention through application, publication, examination and grant. A middle group of chapters deals with the patent once granted: the rights it confers, the registers in which dealings with it are recorded, and the ways it can be attacked through opposition, revocation and surrender. A further group deals with the patent’s relationship to the public: the principles governing working, the compulsory licence, use by government, and the statement of working that a patentee must file. The machinery provisions, the Patent Office, the Controller, the examiners, fees, penalties for offences such as falsely representing an article as patented, and the appeal routes, occupy the closing portions. The field as a whole, and the way this Act sits beside the trade marks, copyright and designs statutes, is mapped in the guide to intellectual property law in India, which this pillar draws on for the comparative frame rather than repeating it.

The chapter map makes the sequence concrete, and a reader navigating the Act for the first time should hold it. Chapter II holds the exclusions. Chapter III holds applications, and Chapter IV publication and examination. Chapter V holds the opposition proceedings, Chapter VI the anticipation code, and Chapter VII the secrecy provisions. Chapter VIII holds grant and the rights conferred, Chapter IX patents of addition, Chapter X amendments, Chapter XI restoration of lapsed patents, and Chapter XII surrender and revocation. Chapter XIII holds the register, Chapter XIV the Patent Office and its establishment, and Chapter XV the Controller’s general powers. Chapter XVI holds working, compulsory licences and revocation for non-working, Chapter XVII the use of inventions for purposes of government, and Chapter XVIII suits concerning infringement. Chapter XIX held the appeals to the Appellate Board and stands emptied of its original content by that Board’s abolition, its function absorbed by the High Courts; Chapter XX holds the penalties, Chapter XXI the patent agents, Chapter XXII the international arrangements, and Chapter XXIII the miscellaneous provisions, among them the avoidance of restrictive conditions in licence contracts and the fee provisions. The map is worth the space it takes because so much Indian patent argument consists of locating a question in the right chapter: a dispute that belongs to Chapter V is repeatedly argued as if it belonged to Chapter XII, and a working complaint that belongs to Chapter XVI is repeatedly misdescribed as an infringement grievance under Chapter XVIII, and the chapters answer to different actors, different tests and different remedies.

Two features of the architecture repay attention before any detail. The first is that the exclusions stand at the front of the Act, in Chapter II, before the application machinery begins, which is a statement of legislative intent about the order in which questions are to be asked. The second is that the Act treats a patent not as a piece of property that the state registers and then leaves alone, but as a grant held on continuing conditions: the condition that the invention be worked, the condition that its proprietor report on that working, the condition that its price and availability answer to the reasonable requirements of the public, and the condition that the applicant dealt candidly with the Patent Office about what was happening to the same invention abroad. The conditions have teeth, in the form of revocation grounds and compulsory powers, and much of what is distinctive in Indian practice flows from them.

Why does Indian patent law define patentability by exclusion?

Because Parliament wanted the boundary of the monopoly fixed by legislation rather than left to examination practice. Novelty and inventive step are judgments made case by case; a statutory list of what is not an invention removes whole classes of subject matter from contention in advance, so applicants, examiners and competitors know the ground rules before any application is filed.

The gateway: what counts as an invention

Section 2(1)(j) of the Act defines an invention as a new product or process involving an inventive step and capable of industrial application. Each element does work. Newness is measured against what was publicly known or used, in India or elsewhere, before the priority date of the claim, and the Act spells out anticipation in detail: publication anywhere, prior claiming in an Indian application, and public knowledge or public use in India each defeat novelty, subject to a small set of exceptions that include display at an exhibition or the reading of a paper before a learned society within a twelve month grace window, where the statutory conditions for the exception are met and, following the Patents (Amendment) Rules of 2024, a prescribed form substantiates the claim. Capability of industrial application, defined in Section 2(1)(ac), means the invention can be made or used in an industry, a deliberately undemanding threshold that excludes almost nothing except the purely theoretical. The three criteria are cumulative, and the way an application fails differs by criterion in a manner that structures the whole practice: a novelty failure is a factual defeat, settled by a document; an inventive step failure is a judgment, argued through the skilled person; an exclusion failure is legislative, turning on the construction of Section 3, which is why the exclusion cases, and not the novelty cases, are the ones that make the law reports and the politics.

The inventive step element carries the real analytical weight, and its statutory definition, recast by the amendment of 2005 in Section 2(1)(ja), is worth following closely. An inventive step is a feature of an invention that involves technical advance as compared to existing knowledge, or has economic significance, or both, and that makes the invention not obvious to a person skilled in the art. Three features of that formulation matter in practice. First, the alternatives are disjunctive: an invention may qualify through technical advance, through economic significance, or through some combination, so an advance that is commercially transformative without being technically deep is not automatically excluded, and an advance that is technically deep without immediate commercial meaning is equally capable of qualifying. Second, the ultimate test remains non-obviousness judged through the eyes of the person skilled in the art, the notional technician who knows everything in the field and invents nothing, and the Indian courts apply that test through the same structured inquiry used elsewhere: identify the inventive concept, identify what was known, identify the differences, and ask whether those differences would have been obvious to the skilled person or required a degree of invention. Third, the definition directs attention to a feature of the invention, which disciplines the applicant to say what the advance actually consists in, a discipline the Patent Office enforces in examination by requiring the technical problem and the claimed solution to be identified rather than gestured at. The economic significance limb has attracted a criticism that should be recorded because it goes to the standard’s integrity. A invention that is technically trivial but commercially successful, the criticism runs, has been rewarded by the market already and needs no monopoly, and a definition that counts economic significance toward inventiveness risks patenting good marketing. The defence of the limb is that in an economy where much genuine innovation is adaptive, taking known components to price points and conditions the originating systems never addressed, a purely technical test of advance undervalues the step that made the technology usable at all, and that the non-obviousness requirement at the end of the definition still filters the merely sold from the genuinely advanced. The definition as enacted holds both ideas at once, and examination practice, which demands that economic significance be tied to the feature claimed rather than to the product’s sales, is where the compromise is administered.

Novelty itself is codified with a completeness that repays attention, because Sections 29 to 34 of the Act assemble the anticipation code that examiners and courts apply. The governing idea is that the state of the art comprises everything made available to the public, anywhere, before the priority date of the claim, whether by written publication, by oral disclosure or by use. One refinement catches applicants out and deserves its own statement: anticipation by prior claiming. Under Sections 11 and 13, an invention can be defeated by matter claimed in an earlier Indian application that was filed before the priority date but published only afterward, so that the anticipating document was secret on the day the later application was filed and becomes visible only when the earlier application is published in its turn. This whole-contents approach to conflicting applications is the reason a freedom-to-operate search in India is never finished at filing: the eighteen month publication lag conceals a stratum of earlier filings whose claims may yet be published over the top of a later applicant’s specification, and the prudent applicant re-searches as the stratum surfaces. The same structure explains the divisional strategy noted above, since the applicant whose own earlier application threatens the later one has a statutory route to consolidate the family rather than lose the later filing to it. Section 11 fixes the priority date of each claim by reference to the application from which it descends, so that a claim carried in a provisional specification, a convention application or a divisional application is judged at the date it was first disclosed in the supporting document, and claims can therefore carry different priority dates within one application. Against that baseline the Act sets a series of defined exceptions, disclosures that do not count: matter obtained from the inventor and published without consent, display at an exhibition notified for the purpose and the reading of a paper before a learned society, each within the twelve month window the sections prescribe and each hedged with conditions including prompt filing thereafter; public working of the invention for the purpose of reasonable trial, where the nature of the invention made public trial necessary; and use or publication in the course of the confidential evaluation of the invention by government. The scheme matters practically because the exceptions are the only shelter an inventor’s own disclosures receive: India has no general grace period of the kind some systems allow an inventor for any publication within a year of filing, and an inventor who publishes a paper, markets the product or demonstrates it publicly outside the defined exceptions before filing has, by that act, destroyed the novelty of what would otherwise have been a patentable invention. The single most common avoidable loss in Indian practice is exactly that, which is why competent advisers treat the filing date, and not the laboratory completion date, as the event around which disclosure decisions are organised.

The remaining definitions in Section 2 repay a practitioner’s attention because they are load-bearing in litigation. A patent is defined simply as a patent granted under the Act, and the patentee as the person entered in the register as grantee or proprietor, which is why registration of assignments matters so much to standing in a suit. The true and first inventor, defined in Section 2(1)(y), does not include the first importer of an invention into India or a person to whom an invention is first communicated from outside India, a deliberate narrowing that denies the title of inventor to the channel of importation and reserves it to origination, echoing the Ayyangar suspicion of the importer monopolist. Person interested, the standing formula that recurs in opposition, compulsory licensing and revocation, is defined to include a person engaged in or promoting research in the same field as the invention, a definition courts have read generously enough to admit the public interest litigant with a genuine research or access engagement, while still excluding the mere intermeddler. Assignee, exclusive licensee and legal representative are each defined for the transactional provisions, and the definition of government undertaking and of the appropriate office machinery provisions complete a section that functions, as definition sections in Indian statutes do, as a compressed statement of the Act’s theory: who originates, who owns, who may complain, and through which office each of them must act.

Passing the Section 2 gateway, however, only earns an invention the right to be tested against Section 3, and it is to the exclusions that any competent Indian patent analysis turns next.

The exclusion list: Section 3

Section 3 opens with words that do the excluding before any criteria are reached: the listed matters “are not inventions within the meaning of this Act”. The section does not say that these things are unworthy, or that they are obvious, or that they fail for want of an inventive step. It removes them from the category of invention itself, which is a stronger and blunter legislative technique than the patentability criteria other systems rely on, and it explains why Indian practitioners begin their advice not with the prior art but with the checklist. The clauses run from (a) to (f) and then from (h) to (p); clause (g) does not exist in the enacted text, a small fossil of the section’s amendment history that examiners’ reports occasionally confuse the unwary into searching for.

The exclusions fall into intelligible groups. One group protects the commons of knowledge: discoveries, scientific principles and abstract theories under clause (c), which pairs the mere discovery of a scientific principle and the formulation of an abstract theory with the discovery of any living thing or non-living substance occurring in nature. The principle is that finding out what the world contains is not making something, and that the first person to isolate a substance that nature had already produced has observed rather than invented, though the moment human ingenuity produces something that did not occur in nature, or a new and non-obvious process for obtaining the natural substance, the analysis changes. A second group, in clauses (a) and (b), polices the boundary between the inventive and the unacceptable: inventions that are frivolous or claim something contrary to well established natural laws, and inventions whose primary or intended use or commercial exploitation would be contrary to public order or morality or would cause serious prejudice to human, animal or plant life or health or to the environment. Clause (b) is the provision invoked against inventions whose exploitation is itself the harm, and its environmental limb gives the Patent Office a textual foothold, rarely used but never repealed, to refuse subject matter whose ordinary use damages the environment seriously.

A third group, clauses (e) and (f), attacks disguised triviality in physical form: a substance obtained by a mere admixture resulting only in the aggregation of the properties of its components, and the mere arrangement, rearrangement or duplication of known devices each functioning independently of the others in a known way. These two clauses do much of the day-to-day work of Indian examination in mechanical and formulation cases respectively. A mixture whose properties are merely the sum of what its ingredients would predictably contribute is an aggregation, not an invention; a combination that produces a result greater than the sum, an interaction between the ingredients or an effect none of them produced alone, crosses the line into patentable territory. The drafting response, repeated in specification after specification, is to plead and where possible evidence synergy, because the clause puts the burden of showing more-than-aggregation squarely on the applicant.

Within clauses (e) and (f) there is a worked distinction that examination reports return to constantly, and an illustration fixes it better than the abstract statement. A known analgesic and a known antacid ground together into a single tablet, each relieving exactly the symptom it always relieved and neither affecting the other’s action, is the paradigm of clause (e): the tablet is convenient, but convenience is not invention, and the claim fails as a mere admixture. The same ingredients in a formulation in which one demonstrably protects the other from degradation in storage, so that the combination works measurably better than either could work in sequence, has crossed into synergy, and the claim is then examined on the ordinary criteria like any other. Clause (f) has its own paradigm: known components bolted together, each performing its established function in the established way, produce an arrangement and not an invention even if the assembly is new as an assembly, while components so integrated that the operation of one changes the operation of another produce the functional interrelationship the clause spares. The lesson the two clauses jointly teach drafters is that Indian examination rewards interaction and punishes aggregation, and that the specification which quantifies the interaction, with comparative data rather than adjectives, is the specification that survives.

The fourth group is the mind-and-information group, and it is the group with the largest commercial consequences. Clause (k) excludes a mathematical or business method, a computer programme per se, and algorithms. Clause (l) excludes literary, dramatic, musical and artistic works, cinematographic works and television productions, which belong to copyright’s territory and are excluded here to keep the boundary between the two statutes clean. Clause (m) excludes a mere scheme or rule or method of performing a mental act or a method of playing a game. Clause (n) excludes a presentation of information, the display or arrangement of data as such, however ingenious the arrangement, and clause (o) excludes the topography of integrated circuits, which India protects, as most jurisdictions do, under a separate specialised statute rather than through patents. Taken together these clauses enact a policy that the patent system is for technology: abstraction, expression, information and mental process belong to other regimes or to no regime, and the monopoly is reserved for contributions to the useful arts.

The exclusion checklist

The table below is the artifact Indian practitioners assemble privately and no official source publishes: each excluded category with its statutory language in substance, an illustration of what falls within the exclusion, an illustration of what falls outside it, and the practical drafting response where one exists.

Excluded category	Statutory language in substance	What falls within it	What falls outside it	Drafting response
Frivolous claims and claims contrary to natural law, Section 3(a)	An invention which is frivolous or which claims anything obviously contrary to well established natural laws	A perpetual motion machine claimed to run without energy input	An unconventional machine whose operation follows established physics	Ground the claim in a stated technical principle and measurable result
Public order, morality and serious prejudice, Section 3(b)	Use or commercial exploitation contrary to public order or morality, or causing serious prejudice to human, animal or plant life or health or to the environment	A device designed for a use the law treats as immoral or injurious	A dual-use technology with ordinary lawful applications	Show the primary and intended use is a lawful one
Discoveries and abstract theory, Section 3(c)	The mere discovery of a scientific principle, an abstract theory, or a living thing or non-living substance occurring in nature	The isolation and naming of a naturally occurring micro-organism as such	A genetically engineered micro-organism that does not occur in nature	Claim the engineered product or the new process, not the discovery
New forms of known substances, Section 3(d)	The mere discovery of a new form of a known substance without enhancement of its known efficacy, and the cognate bars on new properties, new uses and mere new uses of known processes	A salt or polymorph of a known drug with no demonstrated improvement in efficacy	A new form shown by evidence to enhance efficacy significantly	Plead and prove the efficacy enhancement with comparative data
Mere admixture, Section 3(e)	A substance obtained by a mere admixture resulting only in the aggregation of the properties of the components	A blend whose effects are exactly the sum of its ingredients’ known effects	A formulation whose components interact to produce an unexpected effect	Evidence synergy with comparative test results in the specification
Mere arrangement of known devices, Section 3(f)	The mere arrangement, rearrangement or duplication of known devices each functioning independently in a known way	Known components placed side by side, each doing only its own known job	Components combined so that each modifies the working of the others	Claim the functional interaction, not the juxtaposition
Agriculture and horticulture methods, Section 3(h)	A method of agriculture or horticulture	A new method of sowing or of cultivating a crop	Machinery used in farming, which is an article rather than a method	Redirect the claim to the apparatus or implement
Treatment methods, Section 3(i)	Any process for the medicinal, surgical, curative, prophylactic, diagnostic or therapeutic treatment of human beings or animals	A new surgical technique or a method of diagnosing disease practised on the body	A new surgical instrument or a diagnostic device	Claim the device or the in vitro product, not the method on the body
Plants, animals and biological processes, Section 3(j)	Plants and animals in whole or in part, including seeds, varieties and species, and essentially biological processes, other than micro-organisms	A plant variety as such, or a process of conventional cross-breeding	A micro-organism produced by technical intervention, or a biotechnological process	Confine claims to the micro-organism or the technical process
Mathematics, business methods, software and algorithms, Section 3(k)	A mathematical or business method or a computer programme per se or algorithms	A billing method as such, or code claimed without technical effect	A programme producing a demonstrated technical effect in a system	Plead the technical effect and tie the claims to it
Works of authorship, Section 3(l)	Literary, dramatic, musical or artistic works, cinematographic works and television productions	A film or a musical composition offered as an invention	A novel apparatus for recording or reproducing works	Route the subject matter to copyright protection instead
Mental acts and games, Section 3(m)	A mere scheme or rule or method of performing a mental act or a method of playing a game	Rules for a card game, or a method of solving puzzles in the head	Equipment whose design embodies a game in physical form	Claim the physical apparatus where one genuinely exists
Presentation of information, Section 3(n)	A presentation of information as such	A new layout for displaying data, claimed as the display itself	A display technology that changes how a device functions	Claim the functional advance in the device, not the layout
Integrated circuit topography, Section 3(o)	The topography of integrated circuits	A chip layout claimed under the Patents Act	The layout protected under the semiconductor layout-design statute	File under the specialised statute instead
Traditional knowledge, Section 3(p)	An invention which in effect is traditional knowledge, or an aggregation or duplication of known properties of traditionally known components	A claim to the known wound-healing use of a traditionally used substance as such	A genuinely new application demonstrated by technical evidence	Show the step beyond the documented traditional use
Atomic energy, Section 4	An invention relating to atomic energy falling within the Atomic Energy Act, 1962 is not patentable	An invention within the atomic energy field reserved to the state	Subject matter outside the reserved field	None available within the patent system

Two cautions govern the use of any such checklist. It is a map of where the boundary disputes happen, not a substitute for the dispute; nearly every row conceals a live argument about whether the particular subject matter is the excluded thing or the adjacent permitted thing, and the rows for new forms of substances and for computer programmes are argued over in a volume of litigation out of all proportion to their length. And the exclusions interact: an application that escapes clause (d) may still fail clause (e), and one that escapes clause (k) by pleading a technical effect must still show novelty and inventive step on the ordinary criteria. The checklist tells the practitioner where the fight will be. It does not decide the fight.

Section 3(d): the new forms filter and its place

Clause (d) of Section 3 requires separate treatment even within this pillar, because it is the provision on which India’s reputation in world patent politics largely rests and because readers arrive at this article looking for it. Its text, in substance, excludes the mere discovery of a new form of a known substance which does not result in the enhancement of the known efficacy of that substance; the mere discovery of any new property or new use for a known substance; and the mere use of a known process, machine or apparatus unless the known process results in a new product or employs at least one new reactant. An Explanation added with the clause provides that salts, esters, ethers, polymorphs, metabolites, pure forms, particle sizes, isomers, mixtures of isomers, complexes, combinations and other derivatives of a known substance are to be considered the same substance unless they differ significantly in properties with regard to efficacy. The design is legible once the target is named. A pharmaceutical product’s patent life can be extended far beyond its original term if each successive salt, crystal form, dosage form and use of the same molecule earns a fresh twenty years, a practice the industry calls evergreening and its critics call patent term extension by accumulation. Clause (d) answers that practice with a filter: the new form is the same substance in the statute’s eyes, and it becomes patentable only on proof of enhanced efficacy, which the Supreme Court, in the judgment this clause produced, read as therapeutic efficacy for medicines.

The clause’s three limbs operate as a sequence of sieves, and their internal differences matter in practice. The first limb, on new forms, is the efficacy sieve described above, and it is comparative: the applicant must show enhancement over the known substance, which makes the identity of the known substance and the evidence comparing the two the decisive contest in most pharmaceutical oppositions. The second limb, on new properties and new uses of known substances, is categorical in form, and it is the limb that refuses the second medical use claim familiar in European practice, the discovery that a medicine long used for one condition treats another: in India that discovery, however valuable and however rigorously proved, is not an invention, and the discoverer’s protection must be found, if anywhere, in a new dosage form or formulation that itself survives the first limb. The third limb, on the mere use of a known process, machine or apparatus, contains its own internal exception, preserving patentability where the known process yields a new product or employs at least one new reactant, so that process chemistry retains a route to protection for genuine transformations while the repurposing of apparatus for an obvious new task is refused. Read together, the limbs enact a single idea with unusual completeness: knowledge that a thing does more than was known, without a change in the thing or a demonstrated enhancement of what it does, is discovery, and discovery is free.

The judgment itself, Novartis AG v Union of India, and the reasoning by which the Court construed efficacy, the facts of the Glivec application and the fate of the beta crystalline form of imatinib mesylate, belong to the specialist article on Section 3(d) and the Novartis judgment, which owns that litigation. What belongs here is the clause’s position in the architecture. Clause (d) is not a pharmaceutical provision on its face; it speaks of known substances generally, and it applies to chemistry at large. But its centre of gravity is medicines, because medicines are where the economic value of a new form unaccompanied by new efficacy is highest and where the public cost of an unearned extension is paid in prices. It operates alongside, and must not be confused with, the compulsory licence: clause (d) works before grant, by keeping a category of applications out of the system, while the licence works after grant, on patents that cleared every filter and were then not worked or not afforded. The two provisions are the same policy at two different gates, and understanding the Indian system means seeing that the first gate is meant to do most of the work, so that the second is rarely needed.

Bodies, fields and living matter: clauses (h), (i) and (j)

Clause (h) excludes any method of agriculture or horticulture. The exclusion is of methods, not of machinery, so the improved plough, the seed drill and the drip irrigation assembly are patentable articles while the new rotation or cultivation technique practised with them is not, a line that routes agricultural innovation into the implement trade and leaves farming practice itself in the commons. Clause (i) excludes any process for the medicinal, surgical, curative, prophylactic, diagnostic or therapeutic treatment of human beings, and any similar process for animals to render them free of disease or to increase their economic value or that of their products. The reasons are ethical as much as economic: a method of treatment is a physician’s skill exercised on a patient, and the statute declines to let that skill be fenced off by a monopoly that a doctor could infringe by treating. The drafting consequence is a well developed art of indirection: the new surgical instrument is patentable, the diagnostic device is patentable, the substance for use in treatment is patentable as a product, and only the method practised on the body is refused. The clause’s reach into veterinary practice follows the same logic with an economic edge unique to it: processes to render animals free of disease or to increase their economic value or that of their products are excluded alongside the human treatments, so the new method of improving milk yield by treatment is refused where the apparatus that performs a step in it could be claimed. Applicants and examiners accordingly spend their effort on the border between process and product, a border the courts patrol by asking what the claim, read in substance, monopolises: if the answer is an act a practitioner performs on a body, the clause applies whatever the claim is called, and if the answer is a thing that exists independently of any act of treatment, it does not. Diagnostic methods occupy the hardest ground within the clause, because a test performed on a sample removed from the body, in a laboratory, sits awkwardly between the excluded treatment process and the permitted product, and examiners and applicants argue the location and character of the diagnostic step case by case.

Clause (j) excludes plants and animals in whole or any part thereof, other than micro-organisms, including seeds, varieties and species, and essentially biological processes for the production or propagation of plants and animals. Two ideas are joined here. The first is that living matter at the level of the organism is not invention territory: the plant variety is channelled instead into the sui generis regime of the Protection of Plant Varieties and Farmers’ Rights Act, and the interface between that regime and this exclusion is worked through in the article on the Seeds Act and plant varieties, where the variety question is owned. The second is the micro-organism exception, which is the door through which biotechnology entered Indian patent law: a micro-organism that does not occur in nature as claimed, produced or modified by technical intervention, is outside the exclusion, following the logic that the Calcutta High Court had already accepted in the Dimminaco litigation before the 2002 amendment put the exception beyond argument, while an essentially biological process, conventional breeding by selection and crossing, remains excluded even though a technical process of genetic modification is not. The seeds of disputes in this clause are the words “essentially biological”, which separate the breeder’s art, excluded, from the laboratory’s intervention, permitted, and which Indian examiners apply by asking where the technical contribution actually sits in the claimed process.

How does India treat an invention built on traditional knowledge?

Section 3(p) excludes an invention which in effect is traditional knowledge, or which aggregates or duplicates the known properties of traditionally known components. The words “in effect” do the work: a claim dressed in modern language to a use that traditional practice already knew is refused, and the Patent Office supports the refusal from the Traditional Knowledge Digital Library.

The traditional knowledge clause did not appear from theory. It appeared from losses. In the 1990s, patents were granted abroad on the wound-healing use of turmeric and on a fungicide derived from neem, both properties recorded in Indian traditional practice for centuries, and India had to fight both grants through foreign revocation proceedings, succeeding in having the turmeric patent cancelled by the United States office in 1997 after the Council of Scientific and Industrial Research produced documentary prior art, and the neem patent revoked by the European Patent Office in 2000. The lesson drawn was that traditional knowledge fails foreign novelty searches not because it is new but because it is undocumented in the sources examiners search, and the remedy built in response was documentation: the Traditional Knowledge Digital Library, created in 2001 as a collaboration between the Council of Scientific and Industrial Research and the government department responsible for Indian systems of medicine, transcribed formulations from the classical texts of Ayurveda, Unani and Siddha into a classified database to which foreign patent offices were given access under agreement, so that the knowledge can defeat applications at examination instead of after grant. Clause (p), supplemented by the revocation ground for anticipation by knowledge available within a local or indigenous community, is the domestic statutory counterpart of the same defensive structure, and between the clause, the library and the revocation ground, India built the most developed traditional knowledge defences in any patent statute.

The atomic energy bar: Section 4

Outside Section 3 entirely stands Section 4, which provides that no patent shall be granted in respect of an invention relating to atomic energy falling within sub-section (1) of Section 20 of the Atomic Energy Act, 1962. The placement is deliberate. Section 3 defines non-inventions for everyone; Section 4 reserves a field to the state regardless of inventiveness, reflecting the statutory monopoly the Atomic Energy Act constructs over atomic minerals, prescribed substances and the processes of the nuclear programme. An applicant whose invention falls within the reserved field has no drafting response available inside the patent system, and the same sensitivity produces the secrecy machinery described below, under which applications of defence or atomic relevance can be subjected to directions that suspend their ordinary progress. The field the section reserves is the field of atomic minerals and prescribed substances and the processes of their production, use and disposal, the domain the Atomic Energy Act commits to the Central Government and its licensees; inventions at the periphery of nuclear technology, in general engineering, instrumentation or materials with ordinary industrial uses, stand outside the bar, and the boundary question in any given case is whether the invention relates to atomic energy in the sense the referenced provision fixes, a question the Controller resolves with the government’s view available to the Office.

The software question: reading “per se”

Of all the exclusion clauses, the one whose meaning is made by case law rather than by its own text is the software limb of clause (k). The clause excludes “a computer programme per se”, and everything turns on the two Latin words. Read at their broadest, they could be decorative; read as the Patent Office at times read them, the clause could have swallowed every invention implemented in software, which in an economy of embedded systems, vehicles, medical devices and telecommunications is most of modern technology. The legislative history settles the intention: the words “per se” were added precisely to ensure that genuine inventions were not refused merely because a computer programme was involved in them, a purpose recorded in the proceedings of the joint parliamentary committee that examined the amending bill. The examination practice, however, oscillated for years, through successive versions of the Patent Office’s Guidelines for Examination of Computer Related Inventions, the 2017 version of which settled on the position that the substance of the claim, and not its form, decides the question.

How do the courts read the words “per se” in the software exclusion?

In Ferid Allani v Union of India, the Delhi High Court held in 2019 that the bar in Section 3(k) is on a computer programme per se and not on all inventions based on a computer programme, and that an invention demonstrating a technical effect or a technical contribution is patentable even though it is implemented in software.

The Allani reasoning proceeds from the character of modern invention. Almost every contemporary artefact, from the automobile to the microwave oven to the systems behind artificial intelligence and distributed ledgers, embodies programmes, and it would be retrograde, the Court said, to hold that the involvement of a programme destroys patentability across that whole field. The operative test the Court endorsed is the test the guidelines had been approaching: does the claimed invention, in substance, produce a technical effect or make a technical contribution, improving the functioning of a system or solving a technical problem in a technical way? If it does, the claim is not to a programme as such and the exclusion does not bite; if the claim is in truth to the programme itself, to an algorithm in the abstract, or to a business method wearing technical dress, the exclusion applies with full force. The practical consequence for drafting is a discipline Indian applicants observe as routine: the specification must articulate the technical problem, and the claims must be tied to the technical effect said to solve it, because an examiner applying Allani will look through the form of the claims to their substance and an applicant who cannot say what the technical contribution is has already lost. The neighbouring limbs of clause (k) are policed by the same substance-over-form method. A mathematical method is excluded as such, but mathematics embodied in a technical process, in signal processing, in control systems, in compression, is examined by asking what the mathematics does in the world rather than what it is in the abstract; and a business method is excluded as such, which is why the Indian filings of the platform economy are drafted relentlessly toward the technical machinery, the servers, the protocols, the interfaces, and away from the commercial scheme the machinery implements. The examination reports in these fields are therefore arguments about the location of the contribution, and the applicant who concedes that the contribution is commercial while dressing the claims in apparatus language has misread the clause: after Allani the dress is looked through in both directions, sparing the technical contribution in software form and catching the business method in hardware form. The frequently repeated statement that software cannot be patented in India is, after Allani, simply wrong; the accurate statement is that software as such cannot, and that the line between the two is the most commercially consequential boundary in Indian patent examination, drawn afresh in every contested application.

The road to grant: applications, specifications and disclosure duties

The application machinery begins with Section 6, which fixes who may apply: the true and first inventor, a person to whom the inventor has assigned the invention, and the legal representative of a deceased inventor. An application is made in the prescribed form to the Patent Office under Section 7, and the routes by which an invention reaches that office are several, and choosing among them is the first strategic decision in any filing programme. The ordinary application is filed in India directly, either with a complete specification at once or with a provisional specification followed by a complete one. The convention application, under Section 135, is filed in India within twelve months of a first application in a convention country and claims the priority date of that first application, which matters because novelty and inventive step are judged at the priority date, so the convention route moves the effective date of the Indian application back to the foreign filing. The international application under the Patent Cooperation Treaty designating India is treated by Section 7(1A) as an application under the Act with the international filing date as its date, and it enters the Indian national phase within the period the Rules prescribe from the priority date, thirty-one months. Two derivative routes complete the map: the divisional application under Section 16, by which an application disclosing more than one invention is split so that the divided subject matter proceeds in its own application carrying the parent’s filing date, a route the Delhi High Court confirmed in the Syngenta litigation rests on the disclosure of the parent specification and not only on its claims; and the patent of addition under Sections 54 to 56, by which the same applicant protects an improvement in or modification of an invention already applied for or patented, the addition sharing the main patent’s remaining term and requiring no separate renewal fees, in exchange for which its inventive step is not judged against the main invention it improves.

Both international routes rest on treaty accessions whose dates fix the modern filing system. India adhered to the Paris Convention and to the Patent Cooperation Treaty in December 1998, and the consequences run through the Act’s machinery: the convention priority of Section 133 and the sections around it, the Treaty chapter under which international applications are processed, and the treatment of international publication and international search within Indian examination. The Indian Patent Office itself serves as an International Searching Authority and International Preliminary Examining Authority under the Treaty, one of the offices worldwide entrusted with that function, and the choice of the Indian office in either capacity is one of the qualifications for expedited examination at home. A foreign applicant’s Indian filing is therefore rarely a first filing in substance: it is characteristically a national phase entry carrying a priority claim, a Section 8 family, and a prosecution history in other offices that the Indian examiner will read, which is why the candour duty and the convention timetable, rather than the inventive criteria, generate the first practical problems in most incoming cases.

The specification is the legal heart of the filing. A provisional specification under Section 9 describes the invention in general terms and starts the clock; the complete specification must follow within twelve months, a period that cannot be extended, failing which the application is deemed abandoned. Section 10 then prescribes the content of the complete specification with a rigour that reflects its constitutional function within the patent bargain: it must fully and particularly describe the invention, its operation or use and the method by which it is to be performed; it must disclose the best method of performing the invention known to the applicant; and it must end with claims defining the scope of the protection sought, claims that must be clear, succinct and fairly based on the matter disclosed. The claims, not the description, fix the monopoly, and the specification as a whole is the consideration the public receives for the grant: disclosure in exchange for exclusivity, which is why insufficiency of description is a ground on which the granted patent can later be revoked. Applicants resident in India face one further threshold rule under Section 39: they must either file first in India and allow six weeks to pass, or obtain the Controller’s permission, before filing abroad, a rule breached at the cost of the Indian application being deemed abandoned and any patent granted on it exposed to revocation.

For a defined class of inventions the ordinary route never begins, because Sections 35 to 38 construct a secrecy regime that sits across the application machinery. Where the Controller considers an application relevant for defence purposes, or where the Central Government forms that view, secrecy directions issue: the specification is not published, the application is not processed in the ordinary way, no information about it may be communicated, and the applicant may not file abroad in respect of the same invention without permission while the directions run. The directions are reviewed at intervals, the section prescribes a six month review cycle and revocation when the need passes, and where an applicant has suffered hardship because an invention under secrecy could not be used, the Act provides for compensation assessed against the use the government has made or might have made of it. Contravention of the secrecy provisions is an offence, and an application proceeded with in defiance of them is liable to be treated as abandoned. The regime is small in caseload and absolute in effect, and it belongs in a complete account of the Act because it marks the point at which the patent system’s founding bargain, disclosure in exchange for protection, is suspended by the state in the state’s own interest: the inventor of the most sensitive subject matter receives neither publication nor grant on the ordinary timetable, and the statute says so in terms.

Section 8 imposes the candour obligation that most distinguishes Indian prosecution from that of other major offices. An applicant who is prosecuting an application for the same or substantially the same invention abroad must file, and keep updated, a statement setting out the details of those foreign applications, and must give an undertaking to furnish the Controller with the details of subsequent foreign filings up to the date of grant in India. When the Controller so requires, the applicant must also furnish the examination history abroad: the search and examination reports, the objections raised, the claims allowed or refused, and the outcome of oppositions and litigation there. The duty runs on the prescribed Form 3, and the Patents (Amendment) Rules of 2024 rationalised its rhythm, requiring the updated statement within three months of the first examination report rather than through continuous periodic filings, with the Controller able to draw on international databases and further particulars required only on a specific request. The teeth are in Section 64(1)(m): failure to disclose the Section 8 information, or furnishing information that was false to the applicant’s knowledge, is a ground for revoking the patent, and the same failure is a ground of opposition both before and after grant.

The judicial reading of the Section 8 duty has travelled a full circle that applicants need to understand, because the ground of revocation is only as dangerous as the courts make it. In the first phase, the Delhi High Court in Chemtura Corporation v Union of India treated the duty in 2009 as mandatory in the strictest sense and allowed non-disclosure to undermine the patentee’s interim protection, and the Appellate Board in its early decisions took an equally unforgiving line, treating breach of the duty as visiting the patent regardless of the applicant’s state of mind. In the second phase, a Division Bench of the Delhi High Court in Koninklijke Philips Electronics v Maj. (Retd.) Sukesh Behl rejected automatic revocation: Section 64, the court held, confers a discretion, expressed in the word “may”, and a court exercising that discretion must ask whether the omission was deliberate and whether the information withheld was material to the grant, so that an innocent or clerical failure with no bearing on patentability will not destroy a patent. The resting position is therefore narrower than the text read cold: the duty remains mandatory, its breach remains a ground of opposition and revocation, but the remedy follows the culpability and the materiality of what was hidden. A deliberate suppression of a foreign rejection that would have changed the Indian outcome stands at one pole; a late update of a Form 3 recording an abandoned foreign case stands at the other; and Indian prosecution practice, which is organised around docketing foreign families precisely because of this section, reflects the court’s middle position rather than either pole.

What separates a patent from the approval to sell a medicine?

A patent is a right to exclude others, granted by the Patent Office; it confers no permission to sell anything. Marketing approval is a separate permission, granted under the drugs law on criteria of safety, quality and efficacy, and a product can hold a patent and no approval, or full approval and no patent.

That separation is not pedantry, and the regulatory article in this series, on the Drugs and Cosmetics Act 1940, treats the approval side in full. The separation matters to this statute at exactly one point, the research and regulatory exemption discussed below, because the exemption exists only to let a competitor prepare the regulatory file during the patent term so that approval can take effect when the term expires. A reader who conflates the two permissions will misunderstand both, believing a granted patent authorises a launch it does not authorise, or that a regulatory refusal has something to say about validity, which it does not.

How the claims will be read, if the patent is ever litigated, should govern how they are drafted, and Indian doctrine on construction follows a settled path. The specification is read through the eyes of the person skilled in the art, as a whole, with the claims given their ordinary meaning in the context of the description, neither expanded by reference to what the inventor hoped to cover nor cut down by reference to the single embodiment the drawings show. The Supreme Court’s decision in Biswanath Prasad Radhey Shyam v Hindustan Metal Industries, the foundational Indian authority on validity, framed the companion inquiry into obviousness in terms that examiners and courts still work with: the court asks what the prior art disclosed, what the claimed advance consists in, and whether that advance was obvious to the skilled person, warning against the wisdom of hindsight that makes every invention look inevitable after it has been explained. Sufficiency doctrine runs alongside: the description must enable the skilled person to perform the invention across the breadth claimed without undue experiment, and a specification that claims a field while teaching a corner of it invites revocation for insufficiency and for claims not fairly based, the two grounds on which ambitious drafting is most often punished. The disciplines of Section 10 are therefore not formalities laid over the real law; they are the real law of what the monopoly will turn out to be worth.

Publication, examination and grant

An application does not stay secret. Under Section 11A, the application, with its specification, is published after the expiry of eighteen months from the date of filing or the date of priority, whichever is earlier, unless it has been withdrawn in the window the section allows or is subject to secrecy directions under Section 35, which the Controller issues for inventions relevant for defence and which suspend publication and ordinary processing while they run. Early publication can be requested where speed serves the applicant, and publication is a consequential event in two directions. For the public, it opens the file to inspection and starts the period in which a pre-grant opposition can be filed against the application. For the applicant, Section 11A(7) confers rights from the date of publication as if a patent had been granted on that date, with the important qualification that infringement proceedings cannot actually be instituted until the patent is granted, so the applicant holds a deferred right that crystallises on grant and reaches back to publication.

Publication alone does not move an application toward grant, because the Indian system separates filing from examination and examines only on request. Under Section 11B, the applicant or another person interested must file a request for examination, and the Patents (Amendment) Rules of 2024 shortened the period for that request from forty-eight months to thirty-one months from the priority date or the date of filing, a change that compresses prosecution timelines across the system and that applicants trained on the older period miss at their peril, because an application for which no request is made in time is deemed to have been withdrawn. On a request being filed, the application is referred under Section 12 to an examiner, who searches for anticipation by previous publication and by prior claiming under Section 13 and reports on patentability, on the Section 3 exclusions, on inventive step, on sufficiency and on formal defects. The report reaches the applicant as the first examination report, and Section 21 gives twelve months from that report to put the application in order for grant, by amendment and argument, failing which the application is deemed abandoned. The Controller considers the application under Sections 14 and 15, may require amendment, may refuse, and decides after hearing the applicant who asks to be heard. Grant follows under Section 43 where the application is found in order and no pre-grant opposition succeeds or remains pending, and the patent carries the date of filing of the application as its own, a dating rule that matters because the twenty year term runs from that date rather than from the grant.

The prosecution timeline that results is longer than applicants from faster offices expect, though it has shortened substantially from the years in which the system’s backlog was its defining feature, and any honest account must describe the position structurally rather than by a number that dates: pendency varies by the technical field of the application, by the office at which it is prosecuted, and by the complexity of the objections raised, and a reader planning around an application should work from the Patent Office’s own published disposal figures for the relevant field rather than from any general average. The 2024 Rules also widened the Controller’s general power under Rule 138 to extend periods and condone delay, by up to six months on request in the cases the rule covers, which softens the system’s historic unforgivingness at its procedural edges while leaving the unforgiving periods, the twelve months for the complete specification after a provisional and the twelve months after the first examination report, guarded by their own machinery rather than by routine indulgence. What the statute itself guarantees is not speed but sequence: publication before examination, examination only on request, a reasoned report, an opportunity to amend and to be heard, and a decision that can be carried on appeal.

Within the examination stage the Act distributes power in a way worth noticing. The examiner investigates and reports but does not decide; the Controller decides, and Section 77 arms the Controller for that purpose with specified powers of a civil court, including summoning witnesses, receiving evidence on affidavit and awarding costs, so that the contests the Act stages inside the Patent Office, oppositions above all, are conducted as evidence-based proceedings rather than as correspondence. The applicant’s corresponding protections run through the same chapters: a right to be heard before an adverse order on the matters the sections enumerate, a right to amend the specification under Sections 57 to 59, subject to the discipline that an amendment may not broaden the scope of the claims or introduce matter not in substance disclosed in the original filing, and a right of appeal. Alongside the ordinary queue, the Rules provide expedited examination for defined classes of applicant, including start-ups, small entities, applicants who chose the Indian office as their international searching or preliminary examining authority under the Treaty, and others the Rules enumerate, a recognition built into the procedure itself that prosecution speed has different value to different applicants. Appeals from the Controller’s decisions on the matters listed in Section 117A, which range from refusal and opposition outcomes to compulsory licensing determinations, lay originally to the Appellate Board and lie, since that Board’s abolition, to the High Courts, where several courts assign intellectual property work to dedicated divisions with specialist rosters. The practical consequence is that Indian patent law’s meaning is made in three places at once: in the examination guidelines the Patent Office issues, in the Controller’s reasoned orders in contested cases, and in the High Courts’ appellate and revocation jurisdiction, and a reader who watches only the statute book will miss two thirds of the law in action.

Why does India examine patents instead of merely registering them?

Because the designers wanted the validity filter applied by the state before the monopoly issued, not left to litigation afterward. Substantive examination, in which an examiner searches prior art and tests the application against the criteria and the exclusions, means the patent that emerges has survived a first official scrutiny, and the register records tested rights rather than untested claims.

The alternative, a registration or deposit system under which applications are granted as filed and validity is decided only when a court is asked, was the model of some older European systems and survives in attenuated forms elsewhere, most visibly in utility model regimes. India examined under the 1911 Act and examined more rigorously under the 1970 Act, and the choice fits the statute’s theory at every point: a system that refuses whole categories of subject matter needs an examiner to apply the refusals, a system with a candour duty needs an office equipped to receive and use the foreign file, and a system that prices access to medicines as a public question cannot leave the discovery of a bad patent to the chance of litigation funded by the party it excludes. The cost of the choice is throughput, examined in the assessment below: substantive examination is slow, skilled work, and the patent office’s examiner strength and training pipeline have been the binding constraint on the system’s speed in every decade since 1972. The compromise embedded in the modern procedure, deferred examination on request with expedited tracks for applicants who value speed, is the system’s way of keeping the examined grant while rationing the examination resource toward applications whose owners actually want it spent.

The two oppositions

The opposition architecture is the second of the statute’s distinctive features, and it exists in two forms because Parliament wanted the validity of a patent tested both before the monopoly begins and shortly after it. Under Section 25(1), after an application is published and before a patent is granted, any person may represent in writing against the grant, by way of opposition. The standing rule deserves emphasis: any person, with no requirement to show an interest, so that a patient group, a competitor, a researcher or a public health organisation can all oppose, and the grounds, enumerated in the section, track the grounds of invalidity generally: wrongful obtaining, anticipation by publication in India or elsewhere or by prior claiming, prior public knowledge or use in India, obviousness and absence of inventive step, subject matter that is not an invention or is excluded under Section 3, insufficiency of disclosure, failure of the Section 8 duty, a convention application filed outside its twelve month window, non-disclosure or wrong mention of the source of biological material, and anticipation by the knowledge of a local or indigenous community. The Patents (Amendment) Rules of 2024 added two disciplines to this route: an official fee for filing the representation, and a screening step under which the Controller considers whether a prima facie case exists before the opposition proceeds against the applicant, changes aimed at the objection that the route had been used to hold applications in suspension without a real case behind them. If a prima facie case is made out, the applicant is notified, evidence and hearing follow, and the Controller decides whether the patent is granted, refused, or granted on amendment.

The pre-grant route has decided some of the most consequential applications in the Act’s history, which is the practical answer to the claim that it is a procedural frill. The application for the beta crystalline form of imatinib mesylate was opposed before grant by generic manufacturers and by a cancer patients’ organisation, and the Chennai Patent Office refused it in 2006 on grounds that included Section 3(d), beginning the chain of proceedings that produced the Supreme Court’s construction of that clause. The application for tenofovir disoproxil fumarate, an antiretroviral, was opposed by an Indian manufacturer and by networks of people living with HIV, and the Delhi Patent Office refused it in 2009, on obviousness and Section 3 grounds, keeping a workhorse medicine of the public treatment programmes in generic supply. In both cases the opposition cost the opponents a fraction of what equivalent litigation would have cost, in both cases the decision came before any monopoly had been exercised in the market, and in both cases the applicant’s remedy of appeal was intact and used. Set against them the delay cases, in which representations of little substance held applications in suspension, and the design tension of the route is fully visible: the same openness that let patient groups defeat weak pharmaceutical applications cheaply lets the ill-founded representation slow a sound one, and the 2024 screening reforms are the system’s attempt to keep the first property while taxing the second.

Under Section 25(2), within one year from the date of publication of the grant, any person interested, a narrower standing formula that requires a real interest in the field, may give notice of opposition to the grant on the same family of grounds. The post-grant opposition is decided with the assistance of an Opposition Board, which the Controller constitutes for the purpose and which examines the evidence and reports with recommendations, the Board’s recommendation period having been shortened by the 2024 Rules; the Controller then decides after hearing the parties, and may maintain, amend or revoke the patent. The remedies on either route are therefore calibrated: pre-grant, the application itself is at stake; post-grant, the patent can be amended to cure the defect, as where claims are narrowed to escape the anticipation shown, or revoked where the defect is incurable.

The procedure inside each opposition is built on written evidence. In the pre-grant form, the representation is examined first for the prima facie case the amended Rules require; where the case is admitted, the applicant must answer, evidence is exchanged in the sequence the Rules prescribe, the parties are heard, and the Controller’s order decides whether the application proceeds, proceeds on amended claims, or is refused, with costs in the Controller’s discretion. In the post-grant form, the opponent files a notice with a statement of the grounds and the evidence relied on, the patentee answers with counter-statement and evidence, the Opposition Board considers the completed record and reports, and the Controller decides. Two features of the design are easy to overlook and both matter. The first is that amendment runs through the whole architecture as the middle remedy: an application or patent shown to claim too widely is characteristically cut down rather than destroyed, so oppositions in India frequently end not in refusal or revocation but in a patent narrower than the one applied for, which is one reason the raw count of oppositions filed overstates the number of monopolies prevented. The second is that the pre-grant route, being a representation by any person rather than a suit between parties, costs the opponent little and risks the opponent nothing in costs in the ordinary case, while the information it forces into the open, the applicant’s own submissions distinguishing the prior art, becomes part of the permanent file that later challengers inherit. Opposition files are therefore read in India the way prosecution files are read elsewhere, as the first draft of every subsequent validity argument.

The existence of both routes, followed by revocation in court, is the foundation of the “three bites” objection examined below. It is also, on the opposing account, a deliberate sequencing of scrutiny: the pre-grant stage tests the application cheaply before any monopoly exists and before any infringer can be sued; the post-grant stage gives affected interests one structured year in which the administrative file is still warm; and the court stage remains for the party with the resources and the grievance to litigate validity in earnest. Whether that sequencing is a feature or a defect is the live argument, and it is stated in full after the compulsory licence, because the two debates are one debate about how much public control a granted or impending monopoly should carry.

The granted patent: rights, term and maintenance

On grant, Section 48 confers the rights that are the point of the system. For a product patent, the patentee may exclude all others from making, using, offering for sale, selling or importing the product in India. For a process patent, the patentee may exclude others from using the process and from dealing in the product obtained directly by that process. The rights are negative in form, a right to stop others, which is why the separation between patent and regulatory approval matters: the patent authorises nothing, it forbids. Section 47 makes every grant subject to express conditions, among them that a patented machine, apparatus or article may be made or used by or on behalf of the government merely for its own use, and that a patented medicine or drug may be imported by the government for its own use or for distribution in government dispensaries and hospitals, reservations that sit in the background of every Indian patent and that distinguish it from a simple private monopoly. The Register of Patents under Section 67 records the patent and the dealings in it, and dealings have their own formalities: an assignment must be in writing and is not effective against third parties until registered, under Sections 68 and 69, so that the register, and not private paper, tells the public who owns the right it must respect. Co-ownership is governed by Section 50, under which each co-owner holds an equal undivided share unless an agreement says otherwise, each may work the invention for the owner’s own benefit without accounting to the others, but a licence or an assignment of a share requires the consent of all, a default position that regularly surprises joint developers who assumed that ownership carried a power to license. Voluntary licensing itself is a matter of contract, constrained by Section 140, which renders void restrictive conditions of defined kinds in contracts for the sale, lease or licence of a patented article, including conditions that tie the buyer to acquiring other articles from the seller or prohibit the buyer from using competing articles or processes, so that the patent monopoly cannot be leveraged by contract into adjacent markets the grant itself does not reach.

The term is fixed by Section 53 at twenty years from the date of filing of the application, the standard the 2002 amendment substituted for the older structure in which process patents in food and medicines ran for shorter periods. The term is not automatic in its enjoyment: renewal fees are payable, beginning with the third year, to keep the patent in force, and a patent whose renewal fee is not paid within the prescribed period, including the extension the section allows on payment of an additional fee, ceases to have effect. Cessation is not necessarily the end, because Sections 60 to 62 permit restoration on application made within eighteen months, where the Controller is satisfied that the failure to pay was unintentional and that the application was made without undue delay; a restored patent carries conditions protecting those who began using the invention while it stood lapsed. Infringement, where it happens, is sued under Section 104 in the district court, with the structural rule that where the defendant counterclaims for revocation the suit travels to the High Court, and the reliefs under Section 108 are an injunction, damages or an account of profits at the plaintiff’s election, and delivery up or destruction of infringing articles. A defendant, under Section 107, may take every ground available under Section 64 as a defence, which is the mechanism by which court revocation is usually reached in practice. Infringement itself is assessed by comparing the defendant’s product or process with the claims, and Indian courts, like others, distinguish the essential integers of a claim from its inessential substitutions, so that a defendant who takes the substance of the invention while varying a non-essential feature is not saved by the variation, while a defendant whose product omits an essential integer stands outside the claim however close the commercial resemblance. Process patents carry one procedural advantage that materially affects litigation, and Section 104A states it. Where the patent is for a process for obtaining a product, and the product is new or there is a substantial likelihood that the defendant’s identical product was made by the patented process and the patentee cannot determine the process actually used despite reasonable effort, the burden shifts: the court may direct the defendant to prove that the process used to obtain the identical product is different from the patented process, subject to protection for the defendant’s manufacturing and business secrets. The provision exists because process infringement happens behind factory doors, and it converts the product on the market into the beginning of the patentee’s case rather than the end of it. In practice it is one of the reasons process claims, the only pharmaceutical claims the 1970 Act allowed for its first three decades, were never the consolation prize they were sometimes described as abroad: in Indian litigation a well-drafted process claim over a new product carried a burden-shifting weapon that product claims do not need.

Two statutory limits on the rights deserve to stand beside the general statement. The rights do not extend to acts done for experiment or research, including the instruction of pupils, which Section 47 read with the scheme of the Act shelters; and the exhaustion principle, expressed for importation in Section 107A read with the rights provisions, treats a sale by or with the authority of the patentee as exhausting the right over the article sold, subject to the conditions the sections impose.

One exemption within the rights scheme is important enough to stand on its own, because it determines the timing of generic competition across the whole pharmaceutical field. Section 107A(a) provides that any act of making, constructing, using, selling or importing a patented invention solely for uses reasonably related to the development and submission of information required under any law in force in India, or in a country other than India, that regulates the manufacture, construction, use, sale or importation of any product, shall not be treated as infringement of patent rights. The provision is India’s version of the exemption other systems call the Bolar exemption, after the American litigation that exposed the absence of one there. Its logic is temporal: a generic manufacturer needs years of development and regulatory work before a medicine can be approved for sale, and if none of that work could lawfully begin until the patent expired, the monopoly would in practice extend years beyond its term, because the first lawful competitor could only start on expiry day. Section 107A(a) moves the start line back: the work may be done during the term, at home and for export markets whose regulators require the information, and only the commercial sale must wait for the term, or for the patent’s earlier fall in opposition, revocation or compulsory licensing. Section 107A(b) completes the provision by treating importation of a patented product from a person duly authorised by the patentee to sell or distribute it as no infringement, the statutory foothold for parallel imports. Litigation over the exemption’s breadth, in particular over whether manufacture for export to jurisdictions where no corresponding patent subsists falls within it, has occupied the Delhi High Court, and the working position practitioners take from the cases is that the exemption covers acts tethered to a genuine regulatory submission, not commercial stockpiling dressed as development. The exemption is, in structural terms, the mirror image of the absence of term extension noted below: India shortens the effective monopoly at neither end by grace, but it refuses to let the monopoly grow at the far end by regulatory delay.

Working, and the powers behind the patent

Chapter XVI of the Act, Sections 83 to 94, is where the 1970 philosophy survives most completely, and it opens not with a power but with a statement of principles. Section 83 declares that patents are granted to encourage inventions and to secure that inventions are worked in India on a commercial scale and to the fullest extent reasonably practicable without undue delay; that they are not granted merely to enable patentees to enjoy a monopoly for the importation of the patented article; that they contribute to the promotion of technological innovation and to the transfer and dissemination of technology, to the mutual advantage of producers and users; and that the protection and enforcement of patent rights contribute to public health and nutrition and to the public interest in sectors of vital importance to socio-economic and technological development. These are general principles for the exercise of the powers that follow, and the Controller and the courts cite them when the compulsory provisions are construed, which gives the section a working life well beyond the declaratory.

The working obligation is kept visible by Section 146, under which every patentee and every licensee must furnish a statement of the extent to which the patent has been worked on a commercial scale in India, in the prescribed Form 27. The statement has been simplified by stages, and under the Patents (Amendment) Rules of 2024 it is filed once in every three financial years, within six months of the close of the period, and no longer calls for the revenue and value detail that patentees had long resisted disclosing; it records whether the invention was worked or not worked, and the patentee’s openness to licensing. Practitioners assembling these filings out of a prosecution record that runs to hundreds of pages of office documents in mixed formats will find that an office file viewer for the specification and record papers is the unglamorous tool the exercise actually turns on. The statement is not a formality without consequence: it is the evidentiary base on which the working questions in a compulsory licence application are argued, and a patentee who reports, year upon year in substance, that a medicine is imported in small quantities at a high price has written the first page of the case against itself.

The working statement has its own litigation history, which illustrates how seriously the working principle is policed outside the Patent Office as well as inside it. For years the Form 27 obligation was widely honoured in the breach, with statements filed late, filed in uninformative terms, or not filed, and the penalty provision was not deployed to discipline the default. The legal scholar Shamnad Basheer carried the matter to the Delhi High Court by public interest litigation, placing Patent Office records before the court to show the scale of non-compliance and the poverty of the information the filed statements contained, and the proceedings ended with the government committing to enforce the obligation and to revise the form, a process that ran through the simplifications of 2020 and 2024. The episode matters to a reader of the Act because it locates the working requirement’s real constituency: not the Office alone, but competitors and researchers who mine the statements, and who treated the statement as the public’s instrument for seeing whether the monopoly was being exercised in India at all. A patentee reading the same history will draw the operational conclusion, which is that the working statement is the document most likely to be quoted back, in oppositions, in compulsory licence proceedings and in print, and should be prepared with that audience in mind.

The compulsory licence itself stands in Section 84. At any time after the expiration of three years from the date of grant of a patent, any person interested may apply to the Controller for a licence on any of three grounds: that the reasonable requirements of the public with respect to the patented invention have not been satisfied; that the patented invention is not available to the public at a reasonably affordable price; or that the patented invention is not worked in the territory of India. Before applying, the applicant must ordinarily have sought a voluntary licence from the patentee on reasonable terms and failed to obtain one within a reasonable period, which the section measures at not ordinarily more than six months. The Controller weighs the application against the factors in Section 84(7), which deem the reasonable requirements of the public unsatisfied in a set of defined circumstances, and against the applicant’s ability to work the invention to public advantage and to bear the risk of doing so. Where a licence is granted, Section 90 governs its terms: the royalty and other remuneration must be reasonable having regard to the nature of the invention and the expenditure in making and developing it; the licensee must work the invention to the fullest extent reasonably practicable; the licence is non-exclusive and non-assignable; its term is tied to the remainder of the patent’s term; and the articles made under it are priced, in the section’s expectation, as low as reasonably possible consistently with the patentee deriving reasonable advantage. Two related mechanisms complete the architecture. Section 91 lets the Controller license related patents where an important technical advance is blocked by an earlier patent, and Section 85 permits revocation of the patent itself for non-working, on application made after two years from the grant of the first compulsory licence, where the invention still has not been worked or the public’s reasonable requirements still have not been met.

Two further powers answer different situations. Section 92 empowers the Central Government, where it is satisfied that a national emergency, or extreme urgency, or a case of public non-commercial use exists, to declare by notification that licences should be granted, whereupon the Controller may license applicants without the ordinary three year wait or the full voluntary negotiation discipline. Section 92A, added in 2005 to implement the Doha Declaration’s solution for countries that cannot manufacture medicines, permits a licence for the manufacture and export of a patented pharmaceutical product to a country with insufficient manufacturing capacity facing a public health problem, on the terms the importing country’s own notification to the international system establishes. And behind all of these stands the government use power of Section 100, under which the Central Government and persons authorised by it may make, use, exercise and vend an invention for the purposes of government at any time after an application has been filed or a patent granted, against remuneration settled by agreement or, failing agreement, by the High Court under Section 103, with the acquisition power in Section 102 for the rare case in which government takes the patent itself.

The Section 84 inquiry itself is more structured than its three grounds suggest, and its structure explains the record that follows. Section 84(6) directs the Controller to take into account the nature of the invention, the time that has elapsed since grant and the measures the patentee has already taken to work it or license it, the applicant’s ability to work the invention to the public advantage and the capital and risk the applicant can command, and whether the applicant first made the required effort to obtain a voluntary licence. Section 84(7) then converts the first ground into a set of concrete failures: the reasonable requirements of the public are deemed unsatisfied where an existing trade or industry in India is prejudiced by the patentee’s refusal to license on reasonable terms, where a demand for the patented article is not met to an adequate extent or on reasonable terms, where an export market for the Indian-made article is injured by conditions the patentee imposes, or where commercial activities in India are unfairly prejudiced by conditions attached to purchase, hire or use of the patented product. Procedure follows the same disciplined pattern under Sections 87 and 88: the Controller first forms a prima facie view, the application is advertised, the patentee and any other person may oppose, and the matter is decided after hearing, with the order determining both the grant and, through Section 90, its economics. A compulsory licence application in India is therefore closer to a small trial of the patentee’s Indian business than to an administrative favour, and applicants who treat it as the latter, as the refused applications discussed below show, fail.

One question runs through the working provisions and decided the Natco case in substance, and it should be isolated because applicants still argue it in every working dispute: does importation count as working? The text of Section 84’s third ground speaks of the invention not being worked in the territory of India, and Section 83’s second principle declares that patents are not granted merely to enable the patentee to enjoy a monopoly for the importation of the patented article. Against that language the patentee’s answer is built on trade reality and on the international framework: the TRIPS agreement forbids discrimination by place of production, a medicine made in one country and imported into another is traded legitimately, and a rule that only local manufacture counts would penalise the most efficient supply chain and could leave patients worse supplied, not better. The Indian authorities have not adopted the extreme of either side. In the Natco proceedings the Controller examined what the importation actually delivered: quantities that reached a small fraction of those who needed the medicine, at the price Bayer charged, and held that working in the sense of the Act, read with Section 83, means supply of the Indian market on terms that answer the reasonable requirements of the public, which importation of that scale and price did not achieve. The resting doctrine is therefore functional rather than geographical in form: manufacture in India is the clearest evidence of working, importation is not automatically its opposite, but importation that leaves the market substantially unsupplied fails the ground as surely as a factory standing idle. Patentees organise their Indian supply and their Form 27 statements around precisely that distinction.

The emergency and export powers have a different and emptier record. No licence under Section 92A for export has been granted, although the mechanism was built for exactly the situation of a manufacturing-poor country facing an epidemic, and demands made in public debate during the pandemic years that the government invoke Section 92 for vaccines and antiviral medicines were not taken up by notification, the government preferring voluntary licensing arrangements and direct procurement. The pattern across Sections 84, 92 and 92A is thus uniform and worth stating as a finding about the Indian state rather than about any one provision: Parliament has repeatedly been willing to enact compulsory powers of great breadth, and the executive has repeatedly been unwilling to be the actor who fires them, leaving the single Section 84 grant, obtained by a private applicant through the adjudicatory route, as the sum of the machinery’s output. Writers who describe Indian compulsory licensing as a standing threat and writers who describe it as a dead letter are both describing this gap between enactment and use, from opposite ends.

The compulsory licence record: one grant

The brief for this article required the record to be stated precisely, and the precise record is short. One compulsory licence has been granted under the ordinary provision of Section 84. In March 2012 the Controller granted Natco Pharma Limited a licence under Bayer Corporation’s Indian patent for sorafenib tosylate, a medicine for advanced kidney and liver cancer sold as Nexavar. The patent had been granted in 2008; Natco had sought a voluntary licence and been refused; the three years from grant had run. The Controller found all three grounds made out: the reasonable requirements of the public were not satisfied because Bayer supplied only a small fraction of the patient population; the price, about Rs 2,80,000 for a month of therapy, was not reasonably affordable against the Rs 8,800 a month at which Natco offered to sell; and importation of that scale was not working in the territory of India in the sense the Act intends. The licence carried a royalty of six per cent of net sales, and set the price at which Natco’s product was to be sold. Bayer appealed. The Intellectual Property Appellate Board dismissed the appeal in March 2013 and raised the royalty to seven per cent. The Bombay High Court, in Bayer Corporation v Union of India, declined to interfere in 2014, and the Supreme Court refused special leave to appeal, leaving the licence intact for the remainder of the patent’s term.

No further compulsory licence has been granted under Section 84 since, and the refusals are as instructive as the grant. The Controller refused BDR Pharmaceuticals a licence for dasatinib, a leukaemia medicine, because a prima facie case had not been made out, and refused Lee Pharma a licence for saxagliptin, a diabetes medicine, on a fuller hearing of the grounds; in neither case did the applicant clear the threshold the Natco decision had established. The rarity of the grant is then read in opposite directions, and the Act’s defenders and critics divide precisely here. The access reading is that a power exercised once in the half century since the modern compulsory regime took shape, and refused where its conditions were not proved, cannot seriously be described as an interventionist threat to research incentives; its work is done by its existence, in price negotiations and voluntary licensing decisions that never reach an application, and a deterrent that never has to fire is a deterrent working as designed. The originator reading answers that the very fact pattern of Natco, in which the price of an imported cancer medicine was itself one of the grounds, announces that a successful imported product sold at the price its research programme requires can be licensed away after three years, and that this announcement is priced into every decision to launch, supply or localise in India, whether or not an application follows. Neither claim is testable on the available record, because the counterfactual, what prices and launches would have been without Section 84, cannot be observed; what the record does establish is narrower and is worth stating plainly: the provision is hard to invoke, harder to sustain, and has been sustained exactly once.

Revocation, and the three bites argument

A granted patent remains liable to revocation under Section 64 on petition by a person interested or the Central Government, or on a counterclaim in an infringement suit, on grounds that largely rehearse the grounds of opposition: that the invention was wrongfully obtained, that it is not an invention or is excluded by Section 3, that it was anticipated or obvious, that the specification does not sufficiently describe it or discloses a best method the applicant withheld, that the claims are not clear or not fairly based, that the patent was granted on a false suggestion, that the Section 8 duty was breached, that the source of biological material was not disclosed, or that the invention was anticipated by the knowledge of a local or indigenous community. Section 66 adds a rarely invoked public interest power under which the Central Government, after hearing the patentee, may revoke a patent it considers mischievous to the State or generally prejudicial to the public. Since the abolition of the Intellectual Property Appellate Board by the Tribunals Reforms Act of 2021, revocation petitions and appeals from the Controller lie to the High Courts, several of which work through dedicated intellectual property divisions.

It is at this point that the two industry positions required by this article’s brief can be stated at full strength, because revocation completes the gauntlet. The originator position, put by the research-based industry and its counsel, is that the Indian system gives an opponent three successive bites at the same question: a pre-grant opposition that any person may file and that can hold an application in examination for years while the applicant has no right it can enforce; a post-grant opposition within a year of grant, revisiting substantially the same grounds; and thereafter revocation petitions and defences in every infringement suit, on grounds that again overlap, so that an Indian patent is never finally settled and its proprietor must price that perpetual contestability into decisions about the Indian market. The exclusions deepen the objection: Section 3 removes categories that competitor systems patent, Section 3(d) in particular demands a proof of enhanced efficacy that no other major office requires in that form, and the compulsory licence stands behind everything, so that the Indian grant is, on this account, the most heavily conditioned in the industrialised patent world, with a predictable cost in the timing of product launches and the location of research investment.

The generic and access position, put by the Indian generic industry, patient organisations and public health researchers, answers each element in the same order. Pre-grant opposition is the cheapest validity filter any system offers: it tests an application before a monopoly exists, before prices are set under it and before an infringer can be sued, and it is routinely used by sophisticated originators in other fields as much as by generic firms. The post-grant year and the revocation jurisdiction mirror mechanisms that exist in Europe, where post-grant opposition at the European Patent Office is a standard stage of a patent’s life, so the Indian structure is a difference of degree rather than of kind. The exclusions are not hidden traps but a published legislative choice, made by a Parliament entitled to decide that new forms without new efficacy, methods of treatment, and business methods will not be monopolised in India, and a filer who knows the list from the day of drafting cannot plausibly claim surprise. And the compulsory licence, used once in five decades and refused on later applications, is a strange candidate for the provision deterring the industry, when the same period saw India become the largest supplier of generic medicines by volume in the world, a position built in the space the 1970 Act deliberately created.

A verdict is possible on part of this, and this pillar takes it. On the three bites argument, the access position has the better of the record: the grounds genuinely overlap, but each stage operates under different standing rules, different evidentiary postures and different remedies, and the 2024 tightening of pre-grant opposition, with its fee and its prima facie screening, was a legislative acknowledgment that the first bite had been used dilatorily, answered with a filter rather than with abolition. On the deeper question, whether the conditioning of the Indian grant suppresses innovation directed at India, the honest answer is that the evidence supports a narrower claim than either side makes: the exclusions demonstrably redirect drafting and development effort around themselves, which is their purpose, and the compulsory licence demonstrably does not operate as a routine instrument of price control, which its critics said it would and its supporters sometimes promised it might. The statute should be judged as what it is, a grant system designed to be argued with, whose designers preferred argument at the boundary to monopoly inside it. On the general accuracy of the positions stated in this article, one qualification must be entered in the manner of the series: what is set out here is the general law of the Patents Act, stated for analysis, and it does not substitute for advice on an individual application or dispute, where the claim language and the dates are usually decisive.

Offences, agents and the integrity of the record

A complete account of the statute has to include its penal and professional machinery, because both shape daily practice out of proportion to the space they occupy in commentary. Chapter XX creates the offences: contravention of the secrecy directions is punishable with imprisonment; falsification of entries in the register is an offence; a person who represents an article as patented when it is not, or applies the word patent or the language of the Patent Office to an unpatented article in a way that misleads, commits the false representation offences, which is why Indian packaging carries its patent numbers carefully; unauthorised claims to the title of patent office practice, and refusal or failure to furnish information the Act requires, carry their own penalties. The penalties are modest by the standards of regulatory statutes, and prosecutions are rare, but the offences mark the boundary of the system’s integrity claims: the register must mean what it says, the word patented must mean what it says, and the state’s reserved fields must stay reserved.

Chapter XXI then regulates the profession that operates the system. A register of patent agents is maintained, entry to which requires the qualifications and the examination the Act and Rules prescribe, and a registered agent alone may practise before the Controller in the representative capacity the Act defines, signing and presenting documents, attending hearings and conducting the prosecution and opposition work that fills this article. The agent provisions are the Act’s acknowledgment that patent practice is a technical profession at the junction of science and law, and the examination discipline behind registration is one of the quieter determinants of the quality of Indian drafting, since the specification failures described above, the undisclosed best method, the claim not fairly based, the Section 3 problem pleaded away rather than answered, are disproportionately the failures of specifications drafted without that discipline. The section of this article addressed to the aspirant reader can be reduced to this: the exclusions are the examinable core of the statute, the term and the two oppositions are its most reliable questions, and the route from application to grant, with its dates and its duties, is where prepared candidates separate themselves.

What the Act withholds: the negative space

The exclusions of Section 3 are the famous negative space of Indian patent law, but the statute’s silences extend beyond them, and three absences define the system by contrast with its counterparts abroad. India has no utility model or petty patent: there is no second-tier right for incremental mechanical innovation of the kind several European and Asian systems register quickly and cheaply, a gap that parliamentary committees and law commission discussions have examined over the years without enactment, so that a small improvement in India must either clear the full inventive step standard or go unprotected. India has no patent term extension or restoration for time lost in regulatory review: the twenty year term runs from filing through examination delay and marketing approval delay alike, and a medicine whose approval consumes a decade of the term is left with the decade that remains, a position the originator industry protests and the access position defends on the ground that the term bargain was never promised to be approval-proof. And India has no statutory data exclusivity for pharmaceutical test data: the confidential regulatory file a company submits to the drugs regulator is protected against dishonest commercial use through the law of confidence and the TRIPS obligation the government accepts, but no fixed period of exclusivity was enacted, the question having been examined by an official committee in 2007 and left where it stood. Each absence is a policy decision of the same character as Section 3, made by declining rather than by enacting, and a reader mapping the Indian system should mark the silences on the same page as the list.

Two closing mechanics of the granted patent complete the picture and are asked about constantly. A patentee may surrender the patent at any time under Section 63, by notice to the Controller, who advertises the offer and hears any person interested before accepting, a route used to end the maintenance burden of a right no longer worth its renewal fees and occasionally used strategically inside settlements. And a patent that has ceased for non-payment and is not restored, or that expires at the end of its twenty years, leaves its subject matter free to all comers, which is the moment the disclosure bargain completes itself: the specification, published since the application’s eighteenth month, becomes the public’s unrestricted manual. The stock of expired Indian patents is accordingly a working technical library, and generic entry on expiry day, prepared lawfully under Section 107A long before, is the system’s designed conclusion rather than an accident at its margin.

The institutions

The Act is administered by the Controller General of Patents, Designs and Trade Marks, who heads the wider intellectual property administration, and within it by the Patent Office established under Section 74, whose head office is at Kolkata and which works through branch offices at Chennai, Delhi and Mumbai, jurisdiction over an application being assigned by the applicant’s place of business or the origin of the invention under the Rules. The Controller, a term the Act uses for the Controller General and the officers to whom power is delegated under Section 73, receives and disposes of applications through examiners, whose statutory duty under Section 12 is the search and investigation that produces the examination report. The system is an examining office, not a registering one: grant follows a substantive search and a reasoned report, and the Controller’s decisions on the matters the Act enumerates, from refusal and amendment to opposition and compulsory licensing, are appealable, following the abolition of the Appellate Board, to the High Court. The constitutional position of the whole structure is worth a sentence: Parliament’s power to legislate on patents flows from the Union List entry on “patents, inventions and designs”, which is why a single central statute, administered by a single central office, governs the field without state variation of the kind that marks so much of the rest of this series. Two administrative features determine how the structure is experienced in practice. The first is the examiner cadre: recruitment, training and retention of examiners with scientific and engineering qualifications set the ceiling on everything the Act promises, and the expansion of that cadre and its reorganisation into technical groups is the unglamorous reform on which the pendency story turns. The second is the publication discipline of the Office itself, whose official journal carries the applications published under Section 11A, the grants, the oppositions advertised and the register changes on which third parties rely, so that the journal, and not rumour, is the operative notice of the system’s state at any moment. An office that examines, opposes, publishes and registers within one institutional roof is a deliberate concentration, and the Act’s procedural law, with its hearings, its evidence and its appeals outward to the courts, is best read as the discipline imposed on that concentration.

Beneath the Act runs a second text that practitioners read as closely as the first: the Patents Rules. The Rules of 2003, which replaced the rules made under the Act in 1972, carry the forms, the fees, the periods and the procedural choreography for everything this article describes, and their amendment history is a record of policy moving at subordinate-legislation speed. The fee architecture distinguishes natural persons, small entities and start-ups from other applicants, pricing access to the system as a deliberate instrument of innovation policy; the amendments of 2016 and after built the expedited examination classes and the start-up facilitators scheme; the amendments of 2019 and 2020 recalibrated examination and working-statement obligations; and the amendments of 2024 compressed the examination request period and reformed the opposition and disclosure routines described above. A reader of the bare Act will therefore miss both the operative deadlines and the living policy of the system, and the converse warning also holds: periods stated in commentary written before the 2024 amendments, above all the forty-eight month examination request period that a generation of guides repeats, are wrong, and the first duty of anyone advising on an Indian application is to check the deadline against the Rules as amended rather than against the textbook on the shelf.

The amendment arc in outline

The 1970 Act has been amended three times in substance, and the trail is owned by the article on the Patents (Amendment) Act 2005 and TRIPS, so only the shape belongs here. The amendment of 1999 built the transitional machinery that TRIPS required of India, the mailbox in which product patent applications for medicines and agrochemicals were held unexamined and the exclusive marketing rights that bridged the decade. The amendment of 2002, in force from 2003, conformed definitions, extended a uniform twenty year term, and recast the opposition into its modern pre-grant form. The amendment of 2005, effective from 1 January 2005, omitted Section 5 and restored product patents in all fields, recast Section 3(d) in the efficacy form described above, and added the export licence of Section 92A. Read together, the three amendments moved the statute from the Ayyangar design to the TRIPS design while retaining, at every stage, the exclusions, oppositions and compulsory machinery that the Ayyangar design had created; that retention, not the amendments themselves, is the continuity that defines the modern Act. One transitional safeguard of the 2005 settlement deserves mention in this pillar because it fixed the rights of an entire generation of producers. The mailbox applications, filed during the decade when product patents were unavailable and examined only from 2005, confronted the system with enterprises that had in the intervening years built factories and launched products in good faith under the process-only regime. The settlement protected them: an enterprise that had made significant investment and was producing and marketing the product before 1 January 2005, and that continued to manufacture it, could not be stopped by the product patent granted on the mailbox application, and was liable instead to pay a reasonable royalty to the patentee. The safeguard is a small masterpiece of transition design, converting what could have been a wave of shutdown litigation into a royalty regime, and it states in miniature the principle the whole settlement observed: obligations accepted abroad were honoured, and the disruption at home was priced and paid for rather than denied.

An honest assessment

What works in the Act is the clarity of its commitments. An applicant in India knows, from the face of the statute, which subject matter is closed, which duties will be exacted, which publics may object and on what grounds, and what the state reserves; the uncertainty of Indian practice is located in application, in examination quality and pendency and in the litigation the granted patent may face, rather than in concealed rules, and that is a defensible allocation of uncertainty for a system whose premise is that the monopoly must argue for itself. What does not work is equally locatable. Examination outcomes in the crowded fields vary between offices and between examiners more than a single national system should permit, and the guidelines culture that has grown up to discipline them, the Manual of Patent Office Practice and Procedure and the subject-specific examination guidelines, is soft law doing hard law’s work, revisable by administrative decision and unevenly internalised. The opposition system, for all its cheapness as a filter, still permits determined parties to stretch proceedings across years, and the appellate corrections that once came from a specialised board depend on the variable capacity of High Court benches. The working provisions announce a principle the enforcement record only intermittently honours, as the Form 27 litigation showed. None of these failures is a failure of the 1970 design as such; they are failures of administrative investment in a design that demands a great deal of its administrators, because a system built on examination, opposition and working review is only as good as the examiners, boards and controllers who run it. The fairest single sentence about the Act is that its text has worn better than its throughput, and that every serious reform proposal of the last decade has been about throughput: faster examination, screened oppositions, simplified statements, digitised records, while the text, exclusions and all, has been left deliberately alone by legislatures of every complexion.

Verdict

The Patents Act 1970 rewards being read the way it is organised, from its refusals inward. A reader who begins at Section 3, and understands that the Indian monopoly is bounded in advance by a legislative list of the unpatentable, will understand the rest of the system quickly: the two oppositions exist because a boundary that important will be contested at every stage, the working and licensing provisions exist because the boundary was drawn to serve domestic manufacture and access, and the single compulsory licence in the record is not an anomaly but the intended relationship between a strong filter at the gate and a strong power in reserve. A reader who begins instead with the grant, and treats the exclusions as exceptions, will misjudge both the practice of the Patent Office and the politics of the statute. There is a larger pattern here that this series has found in statute after statute, and the Patents Act may be its purest commercial instance: the operative content of an Indian statute so often sits in its definitions and its exclusions rather than in its grants. Lawyers trained on the grant provisions of the patent system, its rights, its term and its remedies, are trained on the parts of the Act India shares with everyone; the parts where Indian outcomes are actually decided are the parts a grant-centred reading skips, the list of the unpatentable, the duty of candour, the obligation to work, the standing of any person to oppose. A student who masters the grant and neglects the list will predict Indian outcomes wrongly in both directions, expecting patents the system refuses and fearing monopolies the system has already fenced. The statute’s designers, who placed the exclusions in Chapter II before a single application form is mentioned, understood the priority they were setting. The pillar articles that follow take the argument deeper at the two points where it is most alive: the amendment trail owns how the 2005 settlement was reached, and the Section 3(d) article owns the case that tested the settlement’s most famous clause.

Frequently Asked Questions

Q: Which inventions qualify for a patent in India?

An invention qualifies if it is a new product or process, involves an inventive step and is capable of industrial application, the three criteria in Section 2 of the Patents Act 1970. It must also escape Section 3, which declares whole categories of subject matter not to be inventions at all, and Section 4, which bars inventions relating to atomic energy. Novelty is judged against everything publicly known or used, in India or elsewhere, before the priority date of the claim. Qualification is therefore a two-stage inquiry: the invention must first be the kind of thing the statute is willing to call an invention, and must then be new, non-obvious and industrially applicable within that permitted field.

Q: What is not an invention under the Patents Act?

Section 3 lists the excluded categories: frivolous inventions and claims contrary to well established natural laws; inventions whose use would offend public order or morality or cause serious prejudice to life, health or the environment; discoveries of scientific principles, abstract theories and things occurring in nature; new forms of known substances without enhanced efficacy; mere admixtures aggregating known properties; mere arrangements of known devices; methods of agriculture or horticulture; processes for treating human beings or animals; plants and animals other than micro-organisms and essentially biological processes; mathematical and business methods, computer programmes per se and algorithms; works of authorship; methods of performing mental acts or playing games; presentations of information; integrated circuit topography; and traditional knowledge. Section 4 separately bars inventions relating to atomic energy.

Q: How long does a patent last and from when?

A patent lasts twenty years from the date of filing of the application, under Section 53 of the Act. The term runs from filing, not from grant, so years spent in examination are years consumed out of the twenty, and there is no extension or restoration of term for time lost in examination or in regulatory approval. The twenty years are not enjoyed automatically: renewal fees are payable from the third year onward to keep the patent in force, and a patent whose fee is not paid within the prescribed period, including the permitted extension on payment of an additional fee, ceases to have effect, subject to the restoration procedure. Patents of addition are an exception in form only, sharing the remaining term of the main patent they improve.

Q: How does the patent application process work?

The applicant files at the Patent Office, directly in India, as a convention application within twelve months of a first foreign filing, or through the Patent Cooperation Treaty national phase. The filing carries a provisional specification followed by a complete specification within twelve months, or a complete specification at once. The application is published after eighteen months. Nothing is examined unless a request for examination is filed within the period the Rules prescribe, thirty-one months from the priority or filing date following the 2024 amendment of the Rules. An examiner then searches and reports, the applicant answers the objections within twelve months of the first report, the Controller hears and decides, and grant follows if the application is in order and no pre-grant opposition succeeds.

Q: What is pre-grant and post-grant opposition?

India has two opposition routes. Pre-grant opposition, under Section 25(1), may be filed by any person, without showing any interest, after the application is published and before a patent is granted; since the Patents (Amendment) Rules of 2024 it carries an official fee and is screened for a prima facie case before it proceeds. Post-grant opposition, under Section 25(2), may be filed by any person interested within one year from the date of publication of the grant, and is decided with the help of an Opposition Board. The grounds in both routes track the grounds of invalidity: anticipation, obviousness, excluded subject matter, insufficient disclosure, breach of the foreign-filing disclosure duty and anticipation by traditional knowledge among them. Pre-grant, the application is at stake; post-grant, the patent may be maintained, amended or revoked.

Q: When can a compulsory licence be granted?

Under Section 84, at any time after three years from the date of grant of a patent, any person interested may apply to the Controller for a compulsory licence on three grounds: that the reasonable requirements of the public for the patented invention have not been satisfied; that the invention is not available to the public at a reasonably affordable price; or that it is not worked in the territory of India. The applicant must ordinarily first have sought a voluntary licence from the patentee on reasonable terms and failed to obtain one within a reasonable period, measured at not ordinarily more than six months. Separate provisions cover emergencies and public non-commercial use under Section 92, and manufacture for export to countries with insufficient capacity under Section 92A. Only one licence has been granted under the ordinary provision: to Natco for the cancer medicine sorafenib in 2012.

Q: What is the duty to disclose foreign patent applications?

Section 8 requires an applicant who is prosecuting applications abroad for the same or substantially the same invention to file a statement of those foreign applications with the Indian Patent Office, on the prescribed Form 3, to keep it updated, and to undertake to report further foreign filings until the Indian patent is granted. When the Controller requires it, the applicant must also furnish the foreign examination history: search and examination reports, objections, and the fate of the foreign claims. Failure to disclose, or furnishing information known to be false, is a ground of opposition and a ground for revoking the patent under Section 64(1)(m). The Delhi High Court has held that revocation on this ground is discretionary and turns on whether the omission was deliberate and the information material, so an innocent clerical failure does not by itself destroy a patent.

Q: What is the research and regulatory exemption in patent law?

Section 107A(a) provides that making, constructing, using, selling or importing a patented invention solely for uses reasonably related to the development and submission of information required under a law regulating the manufacture, use, sale or importation of a product, in India or in another country, is not infringement. The exemption, known as the Bolar exemption, lets a generic manufacturer complete the development and regulatory work for a medicine during the patent term, so that approval can take effect when the term ends rather than years afterward; commercial sale must still wait. Section 47 separately shelters acts done merely for experiment or research, including teaching. Section 107A(b) treats importation of a patented product from a person authorised by the patentee to sell or distribute it as no infringement, the parallel import rule.

Q: What is an inventive step under the Patents Act?

Section 2(1)(ja) defines inventive step as a feature of an invention that involves technical advance as compared to existing knowledge, or has economic significance, or both, and that makes the invention not obvious to a person skilled in the art. The three elements work together. The applicant must identify the feature in which the advance sits; the advance may be technical, economic, or a combination; and the ultimate question remains obviousness, judged through the notional skilled person who knows the whole field and invents nothing. In examination and in court the inquiry is structured: identify the inventive concept, identify the state of the art, identify the differences, and ask whether the differences would have been obvious or required invention. Hindsight is the standing error the test guards against, because most inventions look inevitable once explained.

Q: What is a provisional specification, and when must the complete specification be filed?

A provisional specification, filed under Section 9, describes the invention in general terms and secures the filing date while the invention is still being developed; it need not contain claims. The complete specification must be filed within twelve months of the provisional filing. That period cannot be extended, and if the complete specification is not filed in time the application is deemed abandoned. The complete specification must then do the full statutory work under Section 10: describe the invention and the best method of performing it fully and particularly, and end with claims defining the scope of protection, fairly based on the disclosure. Filing provisionally is a legitimate strategy for priority, but the twelve month deadline is the least forgiving date in Indian prosecution, and the complete specification cannot add new matter beyond what fairness to the original disclosure allows.

Q: When is a patent application published, and what follows publication?

Under Section 11A, an application is published after eighteen months from its date of filing or its date of priority, whichever is earlier, unless withdrawn in time or held under secrecy directions. Early publication can be requested. Three consequences follow. The file becomes open to public inspection, so competitors and researchers can read the specification. The window for pre-grant opposition opens, and any person may represent against the grant. And the applicant acquires rights from the date of publication as if the patent had been granted on that date, under Section 11A(7), with the qualification that infringement proceedings cannot actually be started until the patent is granted, when the right reaches back to publication. Publication is therefore the moment an application becomes a public fact with legal consequences in both directions.

Q: What is the deadline for requesting examination, and what happens if it is missed?

An Indian application is examined only on request, under Section 11B. The request must be filed within the period prescribed by the Patents Rules from the date of priority or the date of filing: the Patents (Amendment) Rules of 2024 reduced that period from forty-eight months to thirty-one months. If no request is filed within the period, the application is deemed to have been withdrawn, and the filing is lost with it, however meritorious the invention. The trap is real, because much older commentary and many foreign applicants’ docketing systems still carry the forty-eight month figure. Any person interested, and not only the applicant, may file the request, which competitors occasionally do to force an early decision on a blocking application.

Q: What is a convention application, and how does the priority claim work?

A convention application, under Section 135, is an Indian application filed within twelve months of the first application for the same invention in a convention country, a country notified under the Paris Convention arrangements in Chapter XXII of the Act. The Indian application claims the priority date of the first filing, and novelty and inventive step are judged at that earlier date, so disclosures and publications between the foreign filing and the Indian filing do not defeat the Indian application. Where the first filing was made in several countries, multiple priorities can be claimed for different claims. The discipline is strict: miss the twelve months and the priority claim fails, and a convention application filed outside the window is also a ground of opposition and of revocation. Most foreign-originated Indian applications arrive by this route or by the Patent Cooperation Treaty national phase.

Q: What is a divisional application, and when can one be filed?

A divisional application, under Section 16, splits an application that discloses more than one invention so that the divided subject matter proceeds in a separate application of its own. Division may be required by the Controller where the claims relate to more than one invention, or may be made voluntarily by the applicant at any time before grant. The divisional application carries the filing date and priority position of the parent, and its claims must not extend beyond what the parent specification disclosed. The Delhi High Court’s decision in the Syngenta litigation confirmed that a divisional may rest on inventions disclosed in the parent specification and is not confined to subject matter the parent’s claims had covered, a reading the Patents (Amendment) Rules of 2024 then carried into the procedural text. Division is the standard answer to a unity objection.

Q: What is a patent of addition?

A patent of addition, under Sections 54 to 56, protects an improvement in or modification of an invention for which the same applicant has already applied or holds a patent. It is granted only after the main patent is granted, it requires no separate renewal fees, and it runs for the remainder of the main patent’s term, expiring with it. In exchange, Section 56 provides that the validity of a patent of addition cannot be questioned merely because the improvement lacks inventive step when measured against the main invention, a concession that recognises the applicant is building on the applicant’s own work. If the main patent is revoked, the patent of addition may be converted into an independent patent for the residue of its term. The device suits incremental product development, where each generation improves the last and the family is kept under one term and one fee burden.

Q: What rights does the patentee acquire on grant?

Section 48 gives the patentee the right to prevent all others from making, using, offering for sale, selling or importing the patented product in India; where the patent is for a process, the right extends to using the process and to making, using, offering for sale, selling or importing the product obtained directly by that process. The rights are exclusive but negative: they forbid others, they authorise nothing, which is why a patent confers no permission to market a regulated product. Every grant is subject to the conditions in Section 47, including government use merely for its own purposes and the experimental use shelter. The rights can be assigned or licensed, but an assignment must be in writing and registered to bind third parties, and the Register of Patents under Section 67 is the public record of who holds the right.

Q: Can a granted patent be revoked, and on what grounds?

Yes. Under Section 64, a patent may be revoked on the petition of a person interested or of the Central Government, or on a counterclaim in an infringement suit, on grounds that rehearse the whole validity inquiry: the invention was wrongfully obtained from the petitioner; it is not an invention or falls within Section 3; it was anticipated by prior publication, prior claiming, or public knowledge or use in India; it is obvious; the specification does not sufficiently describe the invention or withholds the best method; the claims are unclear or not fairly based on the disclosure; the patent was obtained on a false suggestion; the Section 8 disclosure duty was breached; the source of biological material was not disclosed or was wrongly stated; or the invention was anticipated by the knowledge of a local or indigenous community. Since the Appellate Board’s abolition, petitions lie to the High Courts.

Q: What happens if renewal fees are not paid, and can a lapsed patent be restored?

A patent ceases to have effect if the renewal fee is not paid within the prescribed period; Section 53 allows the period to be extended on request against an additional fee, but an unpaid patent at the end of that indulgence lapses, and third parties become free to use the invention. Lapse is not always final. Sections 60 to 62 permit the patentee, or the patentee’s legal representative, to apply for restoration within eighteen months from the date the patent ceased, and the Controller restores it if satisfied that the failure to pay was unintentional and that there was no undue delay in applying. Restoration is advertised and can be opposed, and a restored patent carries conditions protecting persons who began using the invention, or made serious preparations to use it, while the patent stood lapsed.

Q: What must a patentee report in the statement of working?

Under Section 146, every patentee and every licensee must furnish a statement, in the prescribed Form 27, of the extent to which the patented invention has been worked on a commercial scale in India. Following the Patents (Amendment) Rules of 2024 the statement is filed once in every three financial years, within six months from the close of the period concerned, and the simplified form no longer demands the revenue and value detail the older form required; it records whether the invention was worked or not worked and the patentee’s position on licensing. The statement is not a dead formality. It is the evidence base on which working is argued in compulsory licence proceedings, and a history of statements recording import-only supply at high prices builds the case a Section 84 applicant later presents. False information in the statement is punishable under the penalties chapter of the Act.

Q: When can the government use a patented invention without the patentee’s consent?

The Central Government, and any person it authorises in writing, may make, use, exercise and vend a patented invention for the purposes of government at any time after an application has been filed or a patent granted, under Section 100. The use is not free: the patentee is entitled to remuneration, fixed by agreement or, failing agreement, by the High Court on a reference under Section 103, taking account of the economic value of the use and the research expenditure behind the invention. Section 102 goes further and lets the government acquire the patent itself where public purpose requires, against compensation. Independently, Section 47 makes every grant subject to the condition that the government may use a patented machine or article merely for its own use, and import a patented medicine for its own use or for distribution in government dispensaries and hospitals. These powers stand behind the compulsory licence provisions and are wider, since they are not tied to the three year wait or the Section 84 grounds.