Invention Disclosure & Patent Basics

Last verified: August 16, 2026. This guide explains general principles of US invention disclosure and university patent practice for researchers and the technology transfer offices (TTOs) that work with them. It is not legal advice. Ownership terms, disclosure deadlines, and assignment obligations are set by your specific institution’s intellectual property policy and any funding agreement in place — always confirm the actual terms with your TTO before acting on a publication or conference deadline.

What an invention disclosure is, in one sentence

An invention disclosure is the internal, confidential written report a researcher submits to their university’s technology transfer office describing a new invention — what it is, how it works, who contributed to it, and what public or funded disclosures have already happened — so the TTO can evaluate whether to pursue patent protection before any public disclosure destroys that option. It is not itself a patent application; see Invention Disclosure vs. Patent Application for how the two differ, and Invention Disclosure Form: A Worked, Filled-In Example for what a completed form actually looks like section by section.

This page is the orientation layer above those pages: who actually owns what you invent at a university, why the timing of disclosure relative to publication is the single highest-stakes decision a researcher-inventor makes, and why the list of authors on your paper and the list of inventors on your patent are answering two legally different questions that frequently produce two different lists.

Who owns a university invention: Bayh-Dole, institutional policy, and your own assignment obligation

There is no single federal rule that gives a university automatic ownership of everything its researchers invent. Ownership rests on two separate layers, and researchers regularly conflate them:

1. Bayh-Dole governs the government’s and the university’s rights against each other — not against the inventor directly

The Bayh-Dole Act (35 U.S.C. §§ 200-212, implemented at 37 CFR Part 401) lets a university, as the federal funding recipient (“contractor”), elect to retain title to a “subject invention” — one conceived or first reduced to practice using federal funds — instead of the government keeping it. But the Supreme Court’s Stanford v. Roche (2011) decision held that Bayh-Dole does not, by itself, vest title in the university. Title still starts with the individual inventor under US patent law; the university only ends up owning the invention if the inventor has independently and effectively assigned their rights to it — typically through an employment agreement, faculty handbook provision, or IP policy the researcher agreed to as a condition of employment, enrollment, or use of university resources. That assignment obligation is an institutional-policy and contract matter, not a Bayh-Dole one.

Under 37 CFR 401.14(c)(1), once an inventor disclosures a subject invention in writing to the university’s designated patent personnel, the university generally has two months to disclose it onward to the funding federal agency. The university must then elect in writing whether to retain title, normally within two years of that disclosure to the agency — but if a public disclosure, sale, or public use has already started the one-year statutory bar described below, the agency can shorten that election window to as little as 60 days before the bar expires (37 CFR 401.14(c)(2)). A late public disclosure doesn’t just threaten patent rights — it compresses the university’s own internal decision clock too.

2. Your own institution’s IP policy and any funding-source terms decide the rest

For inventions with no federal funding involved, ownership is governed entirely by the researcher’s own institution’s intellectual property policy, employment/student agreements, and any sponsor or industry-collaboration terms attached to the specific project — not by Bayh-Dole at all. Most research universities require assignment of inventions made within the scope of employment or using “significant” university resources (lab space, equipment, funding, or supervised personnel), but the exact scope, the treatment of students versus staff versus faculty, and any inventor revenue-share formula are institution-specific and can vary meaningfully between institutions. Several states also have statutes (California Labor Code § 2870 is the most cited example) that limit an employer’s claim to inventions an employee developed entirely on their own time, without employer resources, and unrelated to the employer’s business or the employee’s actual work — if this kind of carve-out applies to you, it comes from your state’s law and your institution’s own policy, not from Bayh-Dole. A separate common-law backstop, “shop right,” can give a university a non-exclusive, royalty-free license to use an employee’s invention even without a full assignment — see Canada’s tech-transfer/IP-policy model for a non-US point of comparison on how institutional IP ownership frameworks are structured elsewhere.

The practical takeaway: “does my university own this?” is answered by reading your institution’s actual IP policy and, where federal funds were used, by Bayh-Dole’s disclosure-and-election machinery layered on top of it — not by a general rule. When in doubt, ask your TTO before you assume either way.

The one-year statutory bar: the single fact that matters most for a publishing researcher

Since the America Invents Act (AIA, effective for applications with an effective filing date on or after March 16, 2013), the United States runs on a first-inventor-to-file system with a narrow safety net: under 35 U.S.C. § 102(b)(1), an inventor’s own public disclosure does not count as prior art against their own later US patent application, provided the application is filed within one year of that disclosure. A “disclosure” for this purpose is broad — it includes a journal publication, a preprint, a conference talk or poster, a thesis or dissertation made publicly available, and in some circumstances a sale or offer for sale (the Supreme Court’s Helsinn Healthcare v. Teva (2019) held that even a confidential commercial sale to a third party can trigger the “on sale” bar).

Two things make this grace period much less forgiving than it sounds for an academic inventor:

It is a US-only safety net. Most major foreign patent offices — including the EPO and China’s CNIPA — apply an absolute-novelty standard with no general grace period for the inventor’s own prior disclosure. A disclosure that is completely harmless to your US rights under the one-year grace period can permanently destroy your ability to patent the same invention in Europe or China, immediately, the moment it becomes public. The EPO’s own exception is narrow (evident abuse by a third party, or disclosure at specific recognized international exhibitions) and limited to six months.
The clock starts the moment the information becomes public — not when the paper is formally published. A conference poster, an accepted-manuscript preprint server post, or a public thesis defense can all start the one-year US clock and the immediate foreign bar, even before the journal’s official publication date.

See 35 U.S.C. § 102: Patent Novelty and Invention Disclosure Timing for the full statutory analysis, and Missed the Invention Disclosure Deadline Before Public Disclosure for what actually happens, and what options remain, if this deadline has already passed.

Why disclosing to your TTO before you submit a manuscript is the decision that protects everything else

Because a provisional or non-provisional patent application filed before any public disclosure sidesteps the grace-period question entirely — both domestically and abroad — the single highest-leverage action a researcher-inventor can take is disclosing to the TTO early enough that a patent application (usually a low-cost provisional application) can be filed before the manuscript, poster, preprint, or thesis goes public. Disclosing to the TTO after a paper is already accepted, or after a conference talk has already happened, does not make patent protection impossible — the one-year US grace period may still apply — but it does convert a routine filing into a deadline-driven, foreign-rights-limited one, and it compresses the university’s own Bayh-Dole election timeline where federal funding is involved (see above).

The disclosure-to-filing sequence, against a publication deadline

Invention reaches a disclosable stage. You have enough of a concrete result — typically once you can describe what the invention is, how it works, and how it differs from what’s already known — to write a meaningful invention disclosure, even if data collection is ongoing.
Submit the invention disclosure to your TTO — before you submit the manuscript, post a preprint, or present the work publicly. This is the step researchers most often get backwards, usually because disclosure feels like a formality to do “once the paper is done.” By the time a paper is done, it is frequently already too late to file ahead of publication.
TTO reviews for patentability and commercial interest. The office assesses novelty, non-obviousness, utility, and likely commercial pathway (see Patentability Assessment: How a TTO Evaluates an Invention Disclosure) and typically runs or commissions a prior art search and, where relevant, a freedom-to-operate analysis (a separate question from patentability: whether practicing the invention would infringe someone else’s existing patent, not whether the invention itself is new).
If the TTO decides to proceed, a provisional patent application is filed — ideally before any public disclosure. A provisional application is comparatively fast and inexpensive to prepare because it requires a written description satisfying 35 U.S.C. § 112(a) but no formal claims, and it establishes a priority date. See Provisional Patent Applications: USPTO Requirements, Cost, and the 12-Month Deadline and the worked example for the mechanics.
Now the manuscript can be submitted, the poster presented, or the preprint posted with the priority date already secured — publication no longer threatens the patent.
Within 12 months of the provisional filing (a deadline that cannot be extended), a non-provisional application must be filed to preserve the priority date, typically claiming benefit under 35 U.S.C. § 119(e). See Provisional Patent vs. Utility Patent Application for how that conversion decision plays out, and patent prosecution for what follows once a non-provisional application is under USPTO examination.

If public disclosure happens before step 4 — a paper goes out, a talk is given, a preprint posts — the sequence does not end, but every remaining step now runs against the one-year US clock described above, and foreign rights may already be gone. That is the entire reason this order matters.

Inventorship vs. authorship: two different questions, two different lists

Researchers routinely assume the people listed as authors on the resulting paper will be the same people listed as inventors on the resulting patent. They are answering different legal questions, decided under different rules, and the two lists diverge more often than most first-time inventors expect.

	Patent inventorship	Paper authorship
What determines it	US patent law — specifically conception: who first formed a definite and permanent idea of the complete and operative invention, as claimed. Governed by 35 U.S.C. §§ 100(f), 116.	Editorial/publisher policy — most journals in the biomedical and many other sciences apply some version of the ICMJE authorship criteria (substantial contribution to conception/design or data; drafting or critically revising the manuscript; final approval; accountability for the work).
Who decides	A legal/factual determination, ultimately reviewable by a court or the USPTO — not a matter of courtesy, seniority, or lab convention.	The corresponding author and co-authors, applying the journal’s stated policy; disputes go to the journal/publisher and institution, not a court, in the first instance.
Contribution that qualifies	Conceiving at least one claim’s inventive concept. Running assays, providing routine technical execution, or supervising the lab without contributing to conception typically does not qualify someone as an inventor, even if that work was essential to the paper.	Substantial intellectual contribution to the design, execution, analysis, or writing of the study — routine technical or supervisory work can qualify for authorship (or at least acknowledgment) even without any inventive conception.
Getting it wrong	Deliberately omitting or including an inventor can render the resulting patent invalid or unenforceable. Honest mistakes can be fixed under 35 U.S.C. § 256; see Correcting Inventorship on an Issued Patent.	Governed by journal policy, COPE guidance, and institutional research-integrity procedures — a correction, retraction, or misconduct finding, not a patent invalidity risk.
Typical mismatch	A collaborator who conceived a key inventive step but did not draft or revise the manuscript can be an inventor without being a paper author.	A collaborator who ran characterization assays central to the paper, or supervised the project, can be a paper author (or fairly expect acknowledgment) without qualifying as a patent inventor.

See Authorship vs. Patent Inventorship for a deeper treatment of this distinction, including how CRediT contributor-role statements interact with (but do not substitute for) an inventorship determination.

Patentability, in brief: the four tests an invention must pass

Once an invention is disclosed, a TTO’s patentability assessment is generally organized around four statutory requirements: utility (35 U.S.C. § 101 — the invention must be useful), novelty (§ 102 — it must not already be disclosed in the prior art, subject to the one-year grace period discussed above), non-obviousness (§ 103 — the invention must not be an obvious variation on existing prior art to a person of ordinary skill in the field), and enablement/written description (§ 112 — the application must describe the invention completely enough that someone skilled in the field could make and use it). See Patent Law Requirements: The Four Tests an Invention Must Pass for the full analysis of each, and How to Conduct a Prior Art Search for how the novelty and non-obviousness tests actually get evaluated against the existing literature and patent record.

Patent assignment: how ownership actually transfers on paper

Even where an institution’s IP policy establishes an inventor’s obligation to assign a covered invention, ownership of a specific patent or application does not transfer automatically — it requires an executed assignment document identifying the specific application or patent, the assignor (the inventor(s)) and assignee (typically the university), signed by the inventor. Most universities handle this in two layers: a standing assignment obligation the researcher agrees to at hire, matriculation, or as a condition of using university resources (in an employment agreement, faculty handbook, or student/postdoc IP policy), and a specific, application-level assignment document the TTO has the inventor sign at or near the time of filing. Under 35 U.S.C. § 261 and the recordation provisions at 37 CFR Part 3, an assignment should be recorded with the USPTO — timely recordation protects the assignee’s ownership against a later bona fide purchaser who was unaware of the earlier, unrecorded assignment. For a researcher, the practical implication is simple: signing the assignment document promptly when the TTO requests it is not paperwork friction to defer — an unrecorded or delayed assignment can create real title uncertainty during licensing negotiations later.

Frequently asked questions

Do I own my invention if I made it without any university funding or resources?

Possibly, depending entirely on your institution’s IP policy and, in some states, statutory carve-outs for inventions made entirely on personal time without employer resources (see the ownership section above). This is not a default federal rule — check your specific institution’s policy and, if genuinely uncertain, ask your TTO directly rather than assuming either answer.

What if I already published or presented before disclosing to the TTO?

US patent rights are not automatically lost — the one-year grace period under 35 U.S.C. § 102(b)(1) may still allow a US filing if you act within a year of the disclosure — but foreign rights, particularly in Europe and China, are very likely already gone. See Missed the Invention Disclosure Deadline for the full breakdown of what remains available and what does not.

Is an invention disclosure the same as a patent application?

No. An invention disclosure is an internal report to your TTO used to evaluate whether to pursue patent protection at all; it is not filed with the USPTO and does not establish a priority date. See Invention Disclosure vs. Patent Application for the full comparison.

Can I be a patent inventor if I’m not listed as a paper author, or vice versa?

Yes, and it happens routinely — see the inventorship-vs-authorship comparison above. The two lists are decided under entirely different rules and commonly diverge.

Does federal grant funding automatically mean the government owns my invention?

No. Under Bayh-Dole, the university (as the funding recipient) can generally elect to retain title to a federally funded “subject invention,” subject to the government’s retained rights, including march-in rights in narrow circumstances — see Bayh-Dole March-In Rights for what those rights actually allow and how rarely they have been exercised.