Patentability Criteria in India: A Complete Guide to What Can and Cannot Be Patented

Patentability Criteria in India: What Can and Cannot Be Patented

Every year, thousands of patent applications are filed at the Indian Patent Office — and a significant share are refused not because the underlying idea lacked merit, but because the applicant misunderstood what Indian law actually demands before it will grant exclusive rights over an invention. Patentability in India is not a single test. It is a layered set of requirements, and an application can clear three of them comfortably and still fail on the fourth.

This guide breaks down exactly what those requirements are, how Indian courts and the Patent Office apply them in practice, where most applications run into trouble, and how to build a filing strategy that survives scrutiny — both at examination and later, if a competitor challenges the grant.

The Legal Foundation: Patents Act, 1970

India’s patent regime is governed by the Patents Act, 1970, along with the Patents Rules, 2003. Section 2(1)(j) of the Act defines an “invention” as a new product or process that involves an inventive step and is capable of industrial application. That single definition is the source of the three substantive tests every patent examiner applies. Layered on top of it are Sections 3 and 4, which list categories of subject matter that Parliament has decided fall outside patent protection altogether — regardless of how novel or clever they might be — and Section 10(4), which governs how completely and clearly the invention must be described in the application itself.

Put together, an invention seeking patent protection in India must clear four independent hurdles:

1. It must be novel.
2. It must involve an inventive step.
3. It must be capable of industrial application.
4. It must be fully and sufficiently disclosed in the specification.

And even after clearing all four, it must not fall into any of the excluded categories under Sections 3 and 4. Missing any one of these is enough for the Controller to refuse the application — or, if the patent has already been granted, for a competitor to have it revoked through opposition.

1. Novelty: Has Anyone Done This Before, Anywhere?

Novelty is the most intuitive requirement, but also the one applicants most often underestimate. Under Section 2(1)(l), an invention qualifies as “new” only if it has not been publicly disclosed anywhere in the world — through publication, prior use, or any other means — before the date the complete specification is filed.

India follows the absolute novelty standard. This is stricter than some applicants expect: it doesn’t matter whether the prior disclosure happened in India or abroad, in a scientific journal or a product brochure, or even whether the person who disclosed it was the inventor. If the invention’s essential features were made publicly accessible before the priority date, novelty is defeated.

There is no broad grace period in Indian law. A founder who demos a working prototype at an industry conference, publishes a preprint, or even discusses technical details with investors without appropriate confidentiality safeguards can inadvertently destroy their own novelty. The Act does carve out a handful of narrow exceptions — disclosures made without the inventor’s consent, communications to government departments, display at a government-notified exhibition, papers read before a learned society (provided the application is filed within twelve months), and limited public working for genuine trial purposes. Outside these specific situations, the rule is unforgiving.

The Supreme Court’s reasoning in Bishwanath Prasad Radhey Shyam v. Hindustan Metal Industries remains foundational here: a straightforward adaptation of something already known to the public does not acquire novelty merely because it has been repackaged or re-presented.

Practical takeaway: A thorough prior art search — covering Indian and international patent databases, academic literature, and commercial use — should happen before any public disclosure, not after.

2. Inventive Step: Would a Skilled Person Have Found This Obvious?

Passing the novelty test only proves an invention hasn’t been done before — not that it deserved a patent. That’s where inventive step (often called “non-obviousness” in other jurisdictions) comes in. Section 2(1)(ja) defines it as a feature that demonstrates technical advancement, economic significance, or both, and that would not have been obvious to a person skilled in the relevant field.

Indian courts assess this through the lens of a hypothetical “person skilled in the art” — someone with ordinary competence and common general knowledge of the field, but no inventive spark of their own. The test asks: faced with the same technical problem and the same prior art, would this hypothetical person have arrived at the claimed solution without needing genuine inventive ingenuity?

The Delhi High Court’s Division Bench, in F. Hoffmann-La Roche Ltd. v. Cipla Ltd. (2015), laid out a structured five-step approach that has since become the standard reference point:

1. Identify the person skilled in the relevant art.
2. Identify the inventive concept embodied in the claim.
3. Determine what was common general knowledge at the priority date.
4. Identify the differences between the prior art and the claimed invention.
5. Ask, without the benefit of hindsight, whether those differences would have been obvious to the skilled person.

The “without hindsight” instruction matters enormously in practice. It is easy, looking backward from a finished invention, to convince yourself the solution was obvious all along. Examiners and courts are trained to resist that bias and instead reconstruct what would genuinely have been apparent before the invention existed.

A mere combination of known components, without any inventive contribution, will not clear this bar — the Supreme Court has been explicit that a “mere workshop improvement” is not patentable, however commercially useful it might be.

3. Industrial Applicability: Does It Actually Work in the Real World?

Section 2(1)(ac) requires that an invention be capable of being made or used in some industry — a term interpreted broadly to cover manufacturing, agriculture, IT, healthcare, and services alike. This requirement exists to keep patents tethered to practical, working technology rather than abstract theory.

A mathematical formula sitting on its own has no industrial application. The same formula, embedded in a specific technical process that produces a concrete, real-world result, may well satisfy this requirement. The dividing line is whether the invention does something usable, or merely describes an idea about what might be possible.

This is also where applications quietly fail without applicants realizing it: a specification that names a desirable outcome (“a method for detecting disease X with improved accuracy”) without explaining how that outcome is actually achieved raises problems on industrial applicability and disclosure simultaneously.

4. Sufficient Disclosure: Can Someone Else Actually Build It?

This is the requirement competitor articles on this topic tend to underplay, yet it trips up a meaningful share of applications. Section 10(4) requires the complete specification to:

• Fully and particularly describe the invention, its operation, and the method of performing it;
• Disclose the best method of performing the invention known to the applicant;
• End with claims that clearly define the scope of protection sought; and
• Be accompanied by a technical abstract.

The underlying legal standard is enablement: a person skilled in the field must be able to actually perform the invention from reading the specification alone, without needing to conduct undue experimentation of their own. Patents are a bargain — the state grants a time-limited monopoly in exchange for the applicant teaching the public how the invention works. An application that withholds the “how” while claiming broad protection breaks that bargain, and examiners (and later, opponents) will flag it.

Common disclosure failures include: describing a desired result without a working mechanism, claims drafted more broadly than what the specification actually supports, and omitting a feature the invention critically depends on. Note that a working prototype is not legally required to file — but the written description must be detailed enough that someone else in the field could build one.

What Cannot Be Patented: Sections 3 and 4 Exclusions

Even an invention that is novel, inventive, industrially applicable, and fully disclosed can still be refused if it falls into a category Parliament has expressly excluded. These exclusions reflect public policy, ethical considerations, and India’s developmental priorities — and they apply regardless of how well the invention otherwise performs against the four tests above.

The Software Myth

One of the most persistent misconceptions is that “software cannot be patented in India.” That’s not accurate. Section 3(k) excludes a computer program per se — that is, code with no technical contribution beyond running on ordinary hardware. A computer-related invention that delivers a genuine technical effect — solving a technical problem in a technical way, not merely automating a business process — can clear this exclusion. The Patent Office’s CRI (Computer-Related Inventions) Guidelines set out a structured framework for assessing exactly where that line falls, and the analysis turns on the substance of what the claim actually achieves, not on whether a computer happens to be involved.

A Note on the Nuclear-Energy Exclusion

Section 4 has historically operated as a blanket exclusion for inventions relating to atomic energy. Recent legislative developments have introduced a conditional, discretionary mechanism under which the government may permit patents for certain peaceful-use nuclear inventions in defined circumstances — but this remains narrow, subject to government direction, and applicants in this space should confirm the current operative position before relying on it, since strategic and national-security-sensitive inventions continue to fall outside patent protection.

Patentability at a Glance

Why an Application Can Still Fail After Grant

Clearing examination isn’t the end of the story. Under Section 25 of the Act, any interested party can oppose a patent — before grant (Section 25(1)) or within one year after grant (Section 25(2)) — on grounds that map directly back to the requirements above: prior publication, prior use in India, obviousness, subject matter that isn’t a patentable invention, and insufficient disclosure. A patent that scraped through examination on a thin novelty argument or a vague specification remains vulnerable to a well-resourced competitor challenging it later, and revocation proceedings under Section 64 offer a further route. This is precisely why the strength of the underlying application matters more than simply obtaining the grant certificate.

Common Misconceptions Worth Correcting

• “A confidential pitch to investors won’t affect novelty.” It depends entirely on the facts — what was actually disclosed, whether real confidentiality obligations were in place, and how those terms are documented. An NDA is a risk-mitigation tool, not an automatic shield.
• “I need a working prototype before I can file.” Not true. What the law requires is a specification detailed enough to enable someone skilled in the field to build the invention — not a physical model.
• “A patentability search after filing is good enough.” A search conducted before drafting shapes how the claims and specification are written in the first place, which is far more valuable than a search that only confirms problems after the application is already locked in.
• “If it’s novel, it’s patentable.” Novelty is one of at least four independent tests, plus the subject-matter exclusions. Many genuinely novel inventions fail on inventive step, industrial applicability, or disclosure sufficiency, or are simply excluded outright under Section 3.

Building a Filing Strategy That Holds Up

Given how many independent requirements have to be satisfied simultaneously, the applications that hold up best under examination — and later under opposition, if challenged — tend to share a few habits:

1. Search before you disclose, not after. A comprehensive prior art search across Indian and foreign sources, conducted before any public disclosure or investor conversation, is the single most effective way to protect novelty.
2. Draft claims around a genuine technical contribution. Especially for software, diagnostics, and pharmaceutical inventions, the claim needs to show what changed technically — not just that something new was assembled.
3. Disclose the best method, not a generic outline. Vague specifications invite objections on both industrial applicability and Section 10(4) disclosure grounds simultaneously.
4. Anticipate the Section 3/4 objection early. If an invention sits close to a business method, an algorithm, a diagnostic method, or a new form of a known compound, build the technical-effect argument into the specification from the outset rather than trying to retrofit it during prosecution.
5. Treat grant as a milestone, not the finish line. A patent remains open to opposition for a year after grant, and to revocation after that — so the underlying application quality matters well beyond the day the certificate is issued.

Why Should You Involve a Patent Lawyer in Kolkata Early in the Process?

Most patentability disputes trace back to decisions made before anyone filed anything. A patent law in India practitioner reviewing your disclosure before drafting can catch Section 3(k) software exposure, flag a shaky novelty position, and build claims around the real technical contribution instead of the sales pitch. At Kuberaa Associates LLP, we run this check alongside our IP rights advocate in Kolkata practice before a single rupee goes toward formal filing.

Patent work bleeds into other areas more than people expect. A startup splitting IP between co-founders often needs our corporate lawyer in Kolkata team to get assignment agreements right, since a sloppy assignment can later be used to argue about who the “true and first inventor” even was. Licensing income from a granted patent brings GST and income-tax questions too, which our taxation and GST advisory desk handles regularly. If a dispute turns into litigation or a criminal complaint over counterfeiting, our criminal lawyer in Kolkata team steps in; if it’s civil, an infringement suit for damages, say, our civil advocate in Kolkata