Patentability is the gatekeeping concept of patent law. Every decision a client and counsel make about an invention — whether to file an application, how broadly to claim, how much to invest in prosecution, whether to assert a patent in litigation, and whether to challenge a competitor's patent — turns on an assessment of patentability. Getting that assessment right early saves clients from spending tens of thousands of dollars prosecuting claims that cannot survive, and from asserting patents that opposing counsel will invalidate.
The stakes operate in both directions. On the offensive side, a patent is only as valuable as the claims that are actually patentable. Broad claims may be commercially attractive, but if they reach into the prior art they will be rejected during prosecution or invalidated after grant. Narrow claims may survive but offer little protection. Counsel must calibrate claim scope against the patentability requirements to secure enforceable rights of meaningful breadth.
On the defensive side, patentability determines whether a patent asserted against a client can be knocked out. A defendant in an infringement suit routinely argues that the asserted claims were never patentable — that they lack novelty under §102, are obvious under §103, are directed to ineligible subject matter under §101, or fail the §112 disclosure requirements. Because an issued patent is presumed valid, the challenger bears the burden of proving invalidity by clear and convincing evidence in district court, but the underlying analysis is the same patentability inquiry the examiner was supposed to perform.
Patentability also varies by jurisdiction. The statutory structure described here is U.S. law. Other jurisdictions apply analogous but distinct standards — the European Patent Convention, for instance, frames the questions as novelty, inventive step, and industrial applicability, and treats subject-matter exclusions differently. Counsel advising on global portfolios must account for these differences rather than assume the U.S. analysis transfers directly.
How It Works
Patentability in the United States is assessed against four statutory provisions of Title 35 of the U.S. Code. An invention must satisfy all of them; failing any one defeats patentability for the affected claims.
Subject-Matter Eligibility and Utility (§101)
Section 101 permits patents on "any new and useful process, machine, manufacture, or composition of matter." Two questions live here. First, utility: the invention must have a specific, substantial, and credible use. Second, and far more contested, subject-matter eligibility. The Supreme Court has long held that abstract ideas, laws of nature, and natural phenomena are not patent-eligible. Courts evaluate eligibility using the two-step framework drawn from Mayo Collaborative Services v. Prometheus Laboratories (2012) and Alice Corp. v. CLS Bank International (2014): step one asks whether the claim is directed to one of these judicial exceptions; if so, step two asks whether the claim recites an "inventive concept" that transforms the exception into a patent-eligible application. Software and diagnostic-method claims are the most frequent battlegrounds. Eligibility is a threshold filter — an invention can clear §101 and still fail the other requirements.
Novelty (§102)
Section 102 requires that the claimed invention be new. An invention lacks novelty — is "anticipated" — if a single piece of prior art discloses every element of the claim. Prior art includes earlier patents, published applications, printed publications, public uses, and products on sale before the application's effective filing date. The America Invents Act shifted the U.S. system to a first-inventor-to-file framework, so the timing of disclosures relative to the filing date is central to the novelty analysis.
Non-Obviousness (§103)
Even a novel invention is unpatentable if the differences between it and the prior art would have been obvious to a person having ordinary skill in the art at the time of the invention. The Supreme Court in KSR International Co. v. Teleflex Inc. (2007) rejected a rigid test and endorsed a flexible, common-sense approach, holding that combining known elements according to known methods to yield predictable results is generally obvious. Obviousness is frequently the most decisive — and most heavily litigated — patentability requirement, because most inventions are improvements on what came before rather than wholly new concepts.
Disclosure Requirements (§112)
Section 112 governs how the invention must be described and claimed. The written description requirement asks whether the specification shows that the inventor actually possessed the claimed invention. The enablement requirement asks whether the disclosure teaches a skilled person to make and use the full scope of the invention without undue experimentation. The definiteness requirement, in §112(b), asks whether the patent claim particularly points out and distinctly claims the invention, so that its boundaries are clear. A claim that is too broad for its disclosure, or too vague to define its scope, fails patentability under §112.
These requirements are tested first during examination. A USPTO examiner who finds a claim unpatentable issues a rejection in an office action, and the applicant addresses it through argument or amendment during patent prosecution. The same requirements are re-litigated later as validity challenges in court or before the Patent Trial and Appeal Board.
Key Considerations for Law Firms
Assess patentability before filing, not after. A prior-art search and a candid patentability opinion at the outset prevent clients from investing in claims that cannot survive §102 and §103. The cost of a search is small relative to a full prosecution that ends in unallowable claims. Tools that organize patent landscapes — for example PatSnap, which supports patent analytics and landscape mapping — can help counsel scope the relevant art before drafting begins.
Calibrate claim scope to the disclosure. The §112 requirements link claim breadth to what the specification actually teaches. Claims drafted broadly to capture future variants must be supported by a written description and enablement commensurate with that scope. Counsel should draft the specification with the eventual claim breadth in mind, not as an afterthought.
Document the obviousness analysis. Because §103 is the most common ground of rejection and invalidation, building a clear record of how the invention departs from the prior art — unexpected results, solving a long-felt need, teaching away — strengthens both prosecution and later defense of validity. AI tools that perform element-by-element comparison against prior art, such as Patlytics, can support the §102 and §103 analysis with claim charting.
Treat §101 strategically in software and life sciences. For inventions in eligibility-sensitive fields, counsel should anticipate Alice/Mayo challenges from the drafting stage — emphasizing technical improvements, concrete implementations, and specific applications rather than abstract results.
Maintain consistency across the family. Patentability assessments should be coordinated across continuations, divisionals, and foreign counterparts. Statements made during prosecution of one application can affect the patentability and enforceability of related claims.
Limitations and Risks
Patentability is a moving target. Subject-matter eligibility under §101 has been notably unstable, with Federal Circuit decisions and USPTO guidance shifting the boundaries of what qualifies. An eligibility opinion that was sound at filing may be undermined by later case law. Counsel should treat §101 conclusions as time-stamped rather than permanent.
Examination is not a guarantee of validity. An issued patent has cleared the examiner's patentability review, but examiners work under time constraints and may not have located the most relevant prior art. A patent can be invalidated post-grant when a challenger surfaces art the examiner never saw. The presumption of validity is rebuttable, not conclusive.
AI patentability tools assist but do not decide. Software that searches prior art, maps landscapes, or charts claims against references accelerates the analysis but does not render a legal conclusion. The novelty and obviousness determinations require attorney judgment about claim construction, the level of ordinary skill, and what the art teaches. AI-surfaced references must be read and assessed, not accepted at face value, and confidentiality and privilege considerations apply to any invention disclosures processed through third-party platforms.
Burden and forum shift the analysis. The same patentability requirements are applied under different burdens depending on the forum — a preponderance standard in USPTO proceedings and examination, clear and convincing evidence in district court. A claim that survives one forum may fall in another, so a patentability opinion should specify the context it addresses.
Jurisdictional divergence. Conclusions about U.S. patentability do not carry over automatically to other jurisdictions, which apply their own subject-matter exclusions and inventive-step standards. Global filing decisions require jurisdiction-specific analysis.