Quick answer: A patentability search looks forward to determine whether a new invention is novel enough to patent, and it happens before filing. An invalidity search looks backward to determine whether an already-granted patent should have issued, and it happens during litigation, licensing, or a validity challenge. Both rely on prior art analysis, but they serve opposite goals.

Understanding the distinction, and knowing when each belongs in your workflow, is foundational to sound prosecution and litigation strategy. This guide breaks down how each search works, when it matters, and what separates a defensible search from a superficial one.

What is a patentability search?

A patentability search evaluates a proposed invention against existing prior art before an application is filed. Also called a novelty search, its purpose is to determine whether an invention, or specific claims within it, is genuinely new and non-obvious in light of what already exists.

This search usually happens early, during invention disclosure review or claim drafting. A strong patentability search helps a team:

• Assess novelty risk before investing in a full application

• Refine claim scope to navigate around close prior art

• Anticipate the rejections an examiner is likely to raise

• Make an informed go or no-go decision on whether to file

The output shapes strategy. If a search surfaces references that disclose key claim limitations, attorneys can narrow claims, add distinguishing elements, or reconsider the filing, well before an office action forces the issue.

What is an invalidity search?

An invalidity search examines a patent that has already been granted and asks whether prior art exists that should have prevented it. These searches are central to litigation and licensing disputes, where the validity of an asserted patent is often the deciding question.

An invalidity search is typically used to:

• Build a defense against an infringement assertion

• Support a petition for inter partes review (IPR) or post-grant review

• Evaluate the strength of a patent before acquisition or licensing

• Assess validity exposure across a portfolio

The bar here is exacting. Invalidity can rest on a single reference that discloses every claim limitation, known as anticipation, or on a combination of references that together render a claim obvious. Either way, the analysis runs element by element, and the disclosure has to hold up under scrutiny. A near-miss reference that feels relevant rarely survives experienced opposing counsel.

Patentability vs. invalidity search: side-by-side comparison

The two searches share a method but differ in timing, intent, and stakes.

In short: a patentability search protects an invention you want to own, while an invalidity search challenges a patent someone else is asserting. The analytical discipline is the same. Only the direction changes.

What makes a prior art search defensible?

Whichever direction you are searching, the difference between a useful result and a hollow one comes down to rigor. A defensible search shares four traits.

It works at the claim element level. A claim is not a single idea. It is a set of limitations that must each be accounted for. Strong searches break claims into their elements and map references against each one, rather than labeling a reference "relevant" at a high level.

It covers patent and non-patent literature. Critical prior art often lives outside patent databases, in technical papers, standards documents, product manuals, and conference proceedings. A search confined to patents alone leaves blind spots that opposing counsel or an examiner will find.

It produces verifiable, cited evidence. Every assertion should trace back to a specific source. When a reference is said to disclose a limitation, a reviewer needs to see exactly where, with a citation they can open and confirm.

It distinguishes strength of disclosure. Not every reference that touches a claim element discloses it fully. The best searches make clear which limitations are squarely disclosed, which are partial, and which are absent, so attorneys can weigh the analysis rather than take it on faith.

A search that meets this standard does more than return a list of references. It produces an evidence-backed picture an attorney can act on and defend.

Why keyword search falls short

Traditional keyword search has a structural limitation: it matches terms, not meaning. Prior art that describes the same concept in different language slips through, and important references that use unfamiliar terminology never surface. For claim work, where the substance of a limitation matters more than its phrasing, that gap is costly.

This is where claim-driven, semantically aware search changes the workflow. Instead of starting from a keyword list, an attorney can search directly from a claim and let the system analyze its language, evaluate disclosures against individual claim limitations, and retrieve relevant references across both patent and non-patent literature. The result is structured analysis with element-level mapping, the way an attorney would assess limitations, rather than a flat ranked list to wade through.

This is the approach behind Leo, Invention Co's AI engine for patentability and invalidity search. Leo begins a search directly from a patent or claim with no keyword construction required, searches patent and non-patent literature simultaneously, and evaluates disclosures against individual claim limitations. It produces element-level mapping between claims and prior art disclosures, organized into structured reports comparable to professional search firm deliverables. For invalidity work, results are benchmarked against references used in real IPR proceedings. The aim is not to replace attorney judgment. It is to move from search to evidence-backed analysis faster, with the rigor that high-stakes prosecution and litigation require.

Key takeaways

• A patentability search runs before filing to confirm an invention is novel and worth protecting.

• An invalidity search runs after grant to test whether a patent can withstand a validity challenge.

• Both depend on the same fundamentals: claim-level analysis, broad patent and non-patent literature coverage, and cited evidence a reviewer can verify.

• Keyword search misses prior art that uses different language. Claim-driven, element-level search closes that gap.

Frequently asked questions

Is a patentability search the same as a novelty search?
Yes. The terms are used interchangeably. Both evaluate a proposed invention against prior art before filing to assess whether it is new and non-obvious.

When should you run an invalidity search?
Run an invalidity search when a patent's validity is in question: when defending against an infringement assertion, preparing an IPR or post-grant review petition, or conducting diligence before acquiring or licensing a patent.

Can the same prior art support both a patentability and an invalidity position?
Yes. The same reference can show that a proposed claim lacks novelty and that an issued claim should not have been granted. What changes is the direction of the analysis and the standard the reference must meet.

What is element-level or claim-level analysis?
It is the practice of breaking a patent claim into its individual limitations and evaluating prior art against each one, rather than judging relevance at the level of the whole claim. It is the standard a defensible search has to meet.

This article is for general informational purposes and does not constitute legal advice.