A research team can publish a five-author paper and file a patent naming only two of those five people as inventors — and both lists can be entirely correct. Authorship and patent inventorship are not two views of the same underlying fact; they are two separate legal determinations, governed by different bodies of rules, that happen to attach to the same piece of research. Understanding why they diverge, and what to do when they do, matters for any researcher whose work is heading toward both a manuscript and an invention disclosure.
Two Different Legal Tests, Not Two Views of the Same Contribution
The confusion usually starts from a reasonable assumption: that “who contributed to this work” is a single question with a single answer. It isn’t. Authorship and inventorship each apply their own test to the same set of facts, and the two tests measure different things.
The authorship test: ICMJE’s four criteria
Most journals and funders anchor authorship to the International Committee of Medical Journal Editors (ICMJE) four-part test. To qualify as an author, a person must meet all four of the following: substantial contribution to the conception, design, or data (acquisition, analysis, or interpretation) of the work; drafting the work or revising it critically for important intellectual content; final approval of the version to be published; and agreement to be accountable for all aspects of the work. Someone who satisfies three of the four should typically be acknowledged rather than listed as an author — see CASRAI’s Authorship term for the full criteria and how they interact with the CRediT contributor-roles taxonomy (ANSI/NISO Z39.104-2022), which documents what each named person did but does not itself decide who qualifies as an author.
The inventorship test: conception of the claimed invention
Patent inventorship asks a narrower and more specific question. Under U.S. patent law, an inventor is someone who contributed to the conception of at least one claim in the resulting patent application — not to the project generally, not to the paper describing it, but to the specific claimed subject matter. The Federal Circuit has defined conception as the formation, in the inventor’s mind, of a definite and permanent idea of the complete and operative invention, as it is thereafter reduced to practice. Building a prototype, running the confirming experiments, or drafting the application are not, by themselves, acts of conception; they are typically reduction to practice, a legally distinct step. When more than one person is involved, the Federal Circuit’s Pannu factors govern who counts as a joint inventor: each must contribute in a significant manner to conception (or reduction to practice) of a claim, that contribution must not be insignificant measured against the full invention, and it must be more than simply explaining well-known concepts or the state of the art. This is governed by 35 U.S.C. § 116 and USPTO’s own examination guidance at MPEP § 2109.01. CASRAI’s Co-Inventors: The Legal Test for Joint Inventorship guide walks through this test, and the Pannu factors specifically, in more depth; the Inventor dictionary term covers the conception-versus-reduction-to-practice distinction and why it matters for how technology transfer offices (TTOs) name inventors on a disclosure.
The practical upshot: ICMJE authorship is a contribution-based test evaluated across four different kinds of involvement in the work as a whole. Patent inventorship is a conception-based test evaluated claim by claim, against the specific language a patent examiner ultimately allows. Two tests, two different units of analysis — which is exactly why they don’t have to produce the same list of names.
Why Someone Can Be a Legitimate Co-Author but Not a Patent Inventor
This is the more common direction of divergence in practice. A person can clearly satisfy all four ICMJE criteria — substantial involvement in the work, drafting or critical revision, final approval, accountability — without having contributed to the conception of the specific claimed invention. Common, entirely legitimate examples:
- A statistician or data analyst who ran the formal analysis, interpreted results, and helped draft the methods and results sections meets the authorship bar through substantial contribution and drafting/revision. But if the underlying inventive concept (a novel assay design, device, or method) was fully conceived before that analysis began, running confirmatory statistics on someone else’s already-conceived invention is not, by itself, conception.
- A graduate student or postdoc who executed experiments designed by a supervisor, and who then helped write up the results, can be a fully legitimate co-author through their experimental and drafting contributions, while the inventive concept they were executing against had already been conceived by someone else before the experiments began.
- A senior collaborator who provided critical intellectual revision of the manuscript — reframing the argument, sharpening the discussion, ensuring the work meets the journal’s standards — satisfies ICMJE’s drafting/revision criterion without ever having touched the invention’s conception.
None of this is authorship inflation or misconduct. It reflects the fact that writing a rigorous, publishable account of a study draws on skills — statistical judgment, experimental execution, scientific writing — that are real, substantial, and authorship-qualifying, but distinct from conceiving a specific patentable idea.
Why Someone Can Be a Patent Inventor but Not a Co-Author
The reverse happens too, and it surprises researchers more often. A person can contribute the key inventive concept — the specific idea that ends up as an allowed patent claim — without ever drafting a word of the resulting manuscript, without providing critical revision, or without agreeing to be accountable for the paper as published. Under ICMJE’s criteria, that person may not qualify as an author even though they are, correctly, a named inventor. Examples this shows up in practice:
- A collaborator who proposes the core idea in a conversation, meeting, or email and then hands off the experimental validation and write-up entirely to others may have conceived the claimed subject matter without doing any of the drafting or revision ICMJE requires for authorship.
- An industry partner, engineer, or clinician outside the core research group who suggests the specific mechanism or design that becomes the patent’s claims may never be involved in, or credited on, the resulting academic paper at all — particularly where the paper and the patent application cover different scopes of the same underlying project.
- A principal investigator who conceives the invention but delegates all manuscript drafting to trainees, and does not meet the ICMJE bar for critical revision of that particular paper (uncommon, but not impossible on a large multi-paper project), could in principle be an inventor on the patent without being an author on every paper describing related work.
This is also why inventorship is assessed claim by claim and can shift during prosecution: as claims are added, narrowed, or dropped between a provisional filing and an issued patent, the correct list of named inventors can legitimately change, even though the author list on an already-published paper cannot be retroactively edited the same way. See CASRAI’s Invention term for how a conceived idea becomes a disclosable, and in federally funded research a Bayh-Dole “subject invention.”
CRediT Roles Are Not a Legal Proxy for Either List
It is tempting to treat the CRediT “Conceptualization” role as a stand-in for patent inventorship, since both use the word “conception” or “conceptualization” in roughly the same sense — forming the initial idea. Resist that shortcut. CRediT is a contribution-documentation taxonomy (ANSI/NISO Z39.104-2022), not a legal standard, and a person can be tagged with the Conceptualization role for contributing to the general research idea or study design without that contribution rising to conception of a specific, claimable invention in the patent-law sense. Similarly, someone can conceive a patentable claim without their CRediT statement using the Conceptualization label at all, if the taxonomy’s role definitions don’t map cleanly onto what they actually did. Treat CRediT and ICMJE authorship criteria as one legal/documentary system, and patent inventorship under 35 U.S.C. § 116 as an entirely separate one that happens to use similar vocabulary.
Practical Guidance for Researchers Navigating Both Processes
- Don’t assume the author list is the inventor list, or vice versa, when preparing an invention disclosure. Most university invention disclosure forms ask specifically who contributed to conception of the claimed idea — not who is a co-author on the associated manuscript. Treat that as a genuinely separate question each time.
- Keep contemporaneous records of when specific ideas were formed. Lab notebooks, dated emails, and meeting notes that corroborate who proposed a specific inventive concept, and when, are what patent counsel and your technology transfer office (TTO) rely on to determine inventorship correctly — author-order negotiations and lab-culture conventions carry no legal weight here.
- Let patent counsel, not the research team, make the final inventorship call. Author order and inclusion are typically settled by consensus among the research group under journal and institutional policy. Inventorship is a legal determination made against the specific claim language a patent examiner allows, and it is normal for a TTO’s patent counsel to reach a different conclusion than the group’s own assumption about “who really did the work.”
- Coordinate publication timing with the patent process early, not after submission. A public disclosure — including a preprint, conference talk, or published paper — can start the clock on patent filing deadlines. See CASRAI’s guides on publication delay and review clauses and on the cost and process of filing a patent for how institutions manage that timing.
- Expect the inventor list to be revisited if claims change during prosecution, even after the associated paper is already published and its author list is fixed. This is normal and does not imply an error in either list.
- Being omitted from one list is not a comment on the value of your work. A contributor who is a strong co-author but not an inventor made a real, authorship-qualifying contribution; a named inventor who isn’t a co-author made a real, claim-qualifying one. Neither outcome diminishes the other.
Frequently Asked Questions
Can the same person be listed differently on a paper and a patent from the same project?
Yes, and it is common. The two lists answer different legal questions — ICMJE’s contribution-based authorship test for the paper, and the conception-based inventorship test under 35 U.S.C. § 116 for the patent — so full overlap, partial overlap, or no overlap between the two lists can all be correct outcomes for the same underlying project.
Does being a corresponding author make someone a patent inventor?
No. The corresponding author role is an editorial and administrative function under a journal’s authorship policy; it has no bearing on whether that person contributed to conception of a patentable claim. Corresponding-author status and inventorship are determined by entirely separate criteria.
Who decides who is an inventor if the research team disagrees?
Patent counsel makes the determination, based on claim language and corroborated evidence of who conceived what, not the research team’s own consensus. This is different from authorship disputes, which CASRAI covers separately in its guidance on ghost, guest, and gift authorship and COPE’s authorship-dispute processes.
Can someone be added as an inventor after a patent application is filed?
Yes, inventorship can be corrected during prosecution as claims are added, amended, or dropped, since inventorship is assessed claim by claim. This is a normal part of patent prosecution and is a separate process from correcting authorship on an already-published paper.
Does authorship on the paper guarantee any rights to the patent?
No. Authorship carries no automatic patent rights. Ownership and inventorship of a patent are governed by inventorship law and by the assignment obligations in an inventor’s employment or funding agreements (including, for federally funded work, the Bayh-Dole Act), not by the author list of any associated publication.
