The Pannu Test Explained: How Courts Determine True Inventorship

Key Definitions

Inventorship

The legal determination of who conceived a claimed invention. Under U.S. patent law, only the actual inventor(s) may be named on a patent. Inventorship is distinct from ownership.

Joint Inventorship

When two or more natural persons each contribute to the conception of at least one claim in a patent. Joint inventors do not need to work together, contribute equally, or conceive every claim.

Conception

The mental formation of a definite and permanent idea of the complete and operative invention. Conception is the touchstone of inventorship under Federal Circuit case law.

35 U.S.C. §116

The patent statute governing joint inventorship. It provides that inventors may apply for a patent jointly even though they did not physically work together, did not make the same type or amount of contribution, or did not each conceive every claim.

35 U.S.C. §256

The statute allowing correction of inventorship errors made without deceptive intent. If an inventor is omitted or incorrectly named, the patent can be corrected rather than invalidated.

Pannu v. Iolab Corp.

The 1998 Federal Circuit decision (155 F.3d 1344) that established the three-factor test for evaluating joint inventorship claims.

Introduction

Inventorship sounds straightforward: the people who invented something should be listed as inventors on the patent. In practice, it is one of the most frequently mishandled aspects of patent law, and the consequences of getting it wrong are severe.

Under 35 U.S.C. §102(f) (pre-AIA) and the general inventorship requirements of §116, naming the wrong inventors, whether by including someone who did not contribute or omitting someone who did, can render the entire patent invalid. While §256 allows correction of inventorship errors made without deceptive intent, relying on after-the-fact correction is risky, expensive, and entirely avoidable.

The Pannu test, established by the Federal Circuit in Pannu v. Iolab Corp. (155 F.3d 1344, Fed. Cir. 1998), provides the legal framework that courts use to evaluate whether a person qualifies as a joint inventor. It has been applied consistently in Federal Circuit decisions for over 25 years and remains the controlling standard in 2026. This guide explains each factor, shows how courts have applied it in recent cases, addresses the evolving question of AI-assisted inventions, and describes how Patent Geyser's Pannu Test module helps you validate inventorship before filing.

The Three Factors of the Pannu Test

To qualify as a joint inventor of a claimed invention, a person must satisfy all three of the following factors. Failing any one factor is dispositive: the person does not qualify as an inventor.

Factor 1: Significant contribution to conception

The person must have contributed in some significant manner to the conception or reduction to practice of the invention.

Conception, not implementation, is what matters. A person who builds a prototype based on someone else's complete design is reducing the invention to practice, which is valuable work, but it does not make them an inventor unless they also contributed to the conception. Conversely, a person who conceives a key aspect of the invention's architecture but never writes a line of code is still an inventor if their conceptual contribution meets all three Pannu factors.

For software inventions, conception typically means defining the technical approach: the algorithm, the system architecture, the data structure, the processing method. Implementing that approach in code is reduction to practice. The person who said “we should use a graph-based dependency resolver to orchestrate cross-service workflows” contributed to conception. The person who coded the implementation based on that specification may or may not have, depending on whether they contributed original technical ideas during the process.

The Federal Circuit has clarified that there is no explicit lower limit on the quantum or quality of inventive contribution required. Joint inventors do not need to make the same type or amount of contribution, and they do not need to conceive every claim in the patent.

Factor 2: Not insignificant in quality

The person's contribution must not be insignificant in quality when measured against the dimension of the full invention.

This is a comparative assessment. The court looks at the contribution in the context of the entire claimed invention, not in isolation. A contribution that sounds meaningful in the abstract may be insignificant relative to the patent's actual scope.

The Federal Circuit illustrated this in HIP, Inc. v. Hormel Foods Corp. (66 F.4th 1346, Fed. Cir. 2023). The patent covered a two-step method for precooking bacon. One putative inventor's contribution was suggesting the use of an infrared oven for the preheating step (as opposed to a microwave oven). The court found this contribution insignificant in quality because infrared preheating was mentioned only once in the specification, appeared only as an alternative embodiment in a single dependent claim, and was peripheral to the core invention.

For software inventors, this means: if your contribution appears only in a minor dependent claim, is described as an alternative approach in a single paragraph, or addresses a peripheral feature rather than the core technical mechanism, it may fail Factor 2 regardless of how technically sophisticated it is.

Factor 3: More than explaining well-known concepts

The person must do more than merely explain to the real inventors well-known concepts and/or the current state of the art.

This factor distinguishes inventors from advisors, consultants, and domain experts who share existing knowledge without contributing original inventive thought. A professor who explains a known machine learning technique to a student, who then applies it in a novel way, has not contributed to the invention. An engineer who suggests a novel adaptation of that technique to solve a problem no one has solved before may have.

In Bio-Rad Laboratories v. ITC (996 F.3d 1301, Fed. Cir. 2021), the Federal Circuit found that two individuals failed Factor 3 because their ideas were presented at a level of generality that could not support inventorship and involved nothing more than elements already present in the published prior art.

For software teams, this means: someone who recommends using a well-known framework, points the team to existing open-source libraries, or suggests applying standard design patterns is not an inventor. Someone who conceives a novel technical approach that goes beyond what was previously known in the field is.

Inventorship Is Claim-by-Claim

A critical and frequently overlooked aspect of inventorship law: inventorship is evaluated on a claim-by-claim basis, but the failure to name an inventor of any single claim invalidates the entire patent.

This means that if Person A conceived the invention described in Claims 1 through 15, and Person B contributed the inventive concept behind Claim 16 (meeting all three Pannu factors for that claim), both Person A and Person B must be named as joint inventors. Omitting Person B invalidates the entire patent, not just Claim 16.

The Federal Circuit confirmed this in Plastipak Packaging, Inc. v. Premium Waters, Inc. (55 F.4th 1332, Fed. Cir. 2022), emphasizing that inventorship is evaluated per claim but that incorrect inventorship on any claim affects the entire patent.

For software inventors with co-founders, collaborators, or team members, this makes the inventorship analysis essential before filing. Every person who may have contributed to the conception of any claim must be evaluated against all three Pannu factors.

The AI Inventorship Question

The intersection of the Pannu test and artificial intelligence has been one of the most actively debated topics in patent law over the past several years. Here is where things stand.

Only natural persons can be inventors

The Federal Circuit held in Thaler v. Vidal (43 F.4th 1207, Fed. Cir. 2022) that an AI system cannot be named as an inventor on a U.S. patent application. The Supreme Court declined to hear the case. This means that regardless of how much an AI system contributes to an invention, only natural persons can be listed as inventors.

The 2024 Guidance (now rescinded)

In February 2024, the USPTO issued guidance on AI-assisted inventions that applied the Pannu factors to determine whether a human had made a “significant contribution” when using AI tools. The guidance stated that each claim must have been invented by at least one named natural person, and it used the Pannu framework to evaluate the human's contribution relative to the AI's output.

The November 2025 Revised Guidance (current)

On November 28, 2025, the USPTO rescinded the February 2024 guidance entirely. The revised guidance clarifies several important points.

The Pannu factors apply only to joint inventorship among multiple natural persons. When a single human uses AI to develop an invention, the Pannu factors do not apply. Instead, the traditional conception test governs: did the natural person form a definite and permanent idea of the complete and operative invention?

AI systems are tools, not collaborators. The USPTO explicitly states that AI systems, including generative AI and other computational models, are instruments used by human inventors, analogous to laboratory equipment or software.

No separate legal standard for AI-assisted inventions. The same inventorship standard applies to all inventions. The inquiry remains whether a natural person conceived the claimed invention. Using AI during development does not change the analysis.

The Pannu factors still govern human-to-human joint inventorship. When multiple natural persons collaborate on an AI-assisted invention, the traditional Pannu test applies among those humans. The use of AI does not change the joint inventorship analysis among human contributors.

What this means for inventors using AI tools

If you use AI tools (including Patent Geyser) in your development process, you must still be the person who conceived the claimed invention. Using AI to generate code, suggest approaches, or draft specifications is analogous to using any other tool. The key question is whether you had the definite and permanent idea of the invention before (or as a result of your interaction with) the AI, or whether the AI independently generated the inventive concept without meaningful human direction. Document your conception process. Record when and how you arrived at the technical solution.

How Patent Geyser's Pannu Test Module Works

Patent Geyser includes a Pannu Test module (part of the Module 4 workflow) that helps inventors validate inventorship before filing. Here is how it works.

The module takes each independent claim from your application and generates three tailored questions, one for each Pannu factor. The questions are not generic. They are dynamically generated based on the specific technical mechanism described in your claim and the white space strategy identified through prior art analysis.

For the Conception factor, the module asks how and when you conceived the specific technical mechanism described in the claim. It probes for timeline evidence, documentation, and the mental process that led to the solution.

For the Quality factor, the module evaluates whether your contribution represents a significant technical advance or merely combines known elements in an obvious way. It assesses technical sophistication and meaningful departure from prior approaches.

For the Known Concepts factor, the module evaluates whether your invention exceeds what was previously known in the field. It explores your awareness of prior art and how your solution differs from existing approaches.

You answer each question in your own words. An AI agent can help you articulate your response if needed, but the substance of the answer must come from you. The module then scores your responses on a 0.0 to 1.0 confidence scale across all three factors and generates a certification status: “Certified” (strong evidence of inventorship), “Needs Clarification” (moderate evidence, additional documentation recommended), or “Rejected” (insufficient evidence).

The output includes a detailed Pannu record text that addresses all three factors with specific technical details from your answers. This record can be used as supporting documentation for inventorship determinations and is included in the project file for practitioner review. This does not replace a legal inventorship determination by a patent attorney. But it structures the analysis, surfaces gaps early, and creates documentation that supports the inventorship record.

Practical Guidance for Getting Inventorship Right

Evaluate inventorship before drafting, not after. Inventorship disputes are far easier to resolve before an application is filed than after a patent issues. Identify all potential contributors early and evaluate each against the Pannu factors for each claim.

Do not conflate contribution with employment. A co-founder, CEO, or project manager is not automatically an inventor. A contractor or consultant might be. Inventorship is based on contribution to conception, not job title, employment status, or equity stake.

Do not conflate implementation with conception. Coding an implementation based on someone else's technical design is not conception. Contributing the original technical idea that the design is based on is. If the coder added novel technical approaches during implementation that appear in the claims, they may be a co-inventor of those claims.

Document conception contemporaneously. Keep records of who contributed what ideas, when, and in what context. Meeting notes, design documents, whiteboard photos, Slack messages, and email threads can all serve as evidence.

Remember that omitting an inventor is worse than including one incorrectly. Both errors can invalidate a patent, but omission is more commonly exploited in litigation. If there is any doubt about whether someone's contribution meets the Pannu test, evaluate it carefully rather than defaulting to exclusion.

Use §256 correction as a safety valve, not a strategy. If an inventorship error is discovered after filing, §256 allows correction without deceptive intent. But correction requires legal proceedings, costs money, and creates a record that can be used against the patent in litigation. Getting it right the first time is always preferable.

Conclusion

The Pannu test is the legal framework that determines who qualifies as an inventor on a U.S. patent. All three factors must be satisfied: significant contribution to conception, contribution that is not insignificant in quality relative to the full invention, and contribution that goes beyond explaining well-known concepts. Inventorship is evaluated per claim, but incorrect inventorship on any claim can invalidate the entire patent.

For software, SaaS, and blockchain teams, where inventions often emerge from collaborative development, getting inventorship right requires deliberate analysis before filing. Every person who may have contributed to the conception of any claim should be evaluated against the Pannu factors, and the results should be documented.

The November 2025 USPTO guidance clarifies that the Pannu factors apply only to joint inventorship among human contributors. AI systems are tools, not inventors. When you use AI in your development process, the inventorship analysis focuses on whether a natural person conceived the claimed invention. Patent Geyser's Pannu Test module structures this analysis for you, but as with all Patent Geyser outputs, the result is a starting point for practitioner review. The PatentFit Directory can help you find a practitioner with experience in your technology area.

Frequently Asked Questions

Understanding the Test

What is the Pannu test?

The Pannu test is the three-factor legal standard from Pannu v. Iolab Corp. (155 F.3d 1344, Fed. Cir. 1998) that courts use to determine whether a person qualifies as a joint inventor under 35 U.S.C. §116. A person must: (1) contribute significantly to conception, (2) make a contribution that is not insignificant in quality relative to the full invention, and (3) do more than merely explain well-known concepts. All three factors must be met. Failure on any single factor means the person does not qualify as a joint inventor.

Can AI be listed as an inventor on a U.S. patent?

No. The Federal Circuit held in Thaler v. Vidal (2022) that only natural persons can be inventors. The USPTO's November 2025 revised guidance reaffirms this, classifying AI systems as tools analogous to laboratory equipment. When a single human uses AI to develop an invention, the traditional conception test applies. When multiple humans collaborate using AI, the Pannu factors govern their joint inventorship, but the AI itself is not part of the analysis.

What happens if inventorship is wrong on a filed patent?

Incorrect inventorship can invalidate the entire patent. Under §256, errors made without deceptive intent can be corrected, but correction requires legal proceedings and creates a record that opponents can use to challenge the patent's validity. The safest approach is to evaluate inventorship carefully before filing using the Pannu factors and to document the analysis.

Applying It

How does Patent Geyser's Pannu Test module work?

The module generates three claim-specific questions, one per Pannu factor, tailored to the technical mechanism in each independent claim. You answer in your own words (with optional AI assistance for articulation). The system scores your responses on a 0.0 to 1.0 confidence scale and produces a certification status (Certified, Needs Clarification, or Rejected) with a detailed record addressing all three factors. This creates structured inventorship documentation for practitioner review before filing.

Should every team member who worked on the software be listed as an inventor?

Not necessarily. Inventorship requires contribution to the conception of at least one claim, not just participation in the project. Coders who implemented someone else's design, project managers who coordinated the work, and testers who validated the output are not inventors unless they contributed original technical ideas that appear in the claims. Evaluate each person against the Pannu factors for the specific claims they may have contributed to.