University Technology Transfer and Bayh-Dole: Who Owns Federally Funded Inventions

Summary. How federally funded research becomes a licensed product, and who owns what along the way.

The problem Bayh-Dole solved

Before 1980, the federal government funded an enormous share of American research and generally took title to whatever inventions resulted. It then licensed them non-exclusively, or not at all.

The results were poor. Companies would not invest the capital required to develop an early-stage invention into a product when a competitor could obtain the same non-exclusive license the day the product succeeded. By the late 1970s the government held roughly 28,000 patents and had licensed under 5% of them.

The Bayh-Dole Act, codified at 35 U.S.C. §§ 200–212, changed the default. Universities, nonprofit research institutions, and small businesses may elect to retain title to inventions made with federal funding, subject to conditions. The stated policy, in 35 U.S.C. § 200, is

"to use the patent system to promote the utilization of inventions arising from federally supported research or development; to encourage maximum participation of small business firms in federally supported research and development efforts; [and] to promote collaboration between commercial concerns and nonprofit organizations, including universities."

The consequence was the creation of an entire institutional apparatus — technology transfer offices, university patent portfolios, licensing programs, and a startup formation pipeline — that did not previously exist. Whether it produced net social benefit is debated. That it produced a functioning system is not.

What is a subject invention

The statute operates on subject inventions, defined in 35 U.S.C. § 201 as any invention of the contractor conceived or first actually reduced to practice in the performance of work under a funding agreement.

Three elements deserve attention.

"Invention" means anything that is or may be patentable. Software, data, and know-how are not subject inventions unless patentable. This matters enormously in practice, because a great deal of valuable university output is copyrightable software or unpatented know-how, governed by institutional policy rather than by Bayh-Dole.

"Conceived or first actually reduced to practice." Either is sufficient. An invention conceived under federal funding is a subject invention even if reduced to practice with private money, and vice versa. Because conception is a mental act that can be difficult to date, the boundary is frequently contested — particularly where a researcher holds concurrent federal and industry funding.

"In the performance of work under a funding agreement." The nexus requirement. Work merely related to the funded project, or performed by the same person on other time, is not automatically covered. Institutions and agencies read this differently, and the safe practice is to disclose and let the determination be made rather than to conclude unilaterally that an invention falls outside.

Mixed funding is the norm, not the exception. A laboratory with a federal grant, an industry sponsor, a foundation award, and institutional funds produces inventions with entangled origins. Contemporaneous laboratory records are the only reliable way to sort them, which is why notebook discipline is a legal requirement and not merely a scientific one.

The deadlines that forfeit rights

Bayh-Dole is a system of conditions, and the conditions are deadlines. The implementing regulations at 37 C.F.R. Part 401 set them out.

Two practical observations. First, the two-month clock starts when the inventor discloses internally, not when the invention is made — which means institutional intake processes determine the deadline. Second, agencies have discretion to extend, and in practice do so with some regularity, but a pattern of late disclosures is a compliance finding that can affect an institution's funding relationships.

Reporting is done through an electronic system, and the administrative burden is real: a research university may file thousands of reports a year.

What the government keeps

Even where the institution retains title, the government retains rights.

The government license. Under 35 U.S.C. § 202(c)(4), the agency receives a non-exclusive, non-transferable, irrevocable, paid-up license to practice the invention throughout the world on behalf of the United States. This license is real and is frequently overlooked in licensing negotiations. A company taking an "exclusive" license to a subject invention is taking a license exclusive of everyone except the United States government, and its agents and contractors performing government work.

March-in rights. 35 U.S.C. § 203 permits the agency to require the contractor to grant a license to a responsible applicant, or to grant one itself, where:

• the contractor has not taken effective steps to achieve practical application within a reasonable time;
• action is necessary to alleviate health or safety needs not reasonably satisfied;
• action is necessary to meet public use requirements specified by federal regulations; or
• the domestic manufacture requirement of § 204 has been breached and no waiver was granted.

No agency has ever successfully exercised march-in rights. Petitions have been filed — most prominently concerning pharmaceutical pricing — and all have been denied, on the reasoning that high price alone does not establish failure to achieve practical application when the product is available. Whether pricing can support march-in has been the subject of proposed framework changes and considerable debate, and it remains the most watched open question in the statute.

The domestic manufacturing preference. 35 U.S.C. § 204 provides that no exclusive right to use or sell a subject invention in the United States may be granted unless the licensee agrees that products embodying the invention will be manufactured substantially in the United States. Waivers are available where domestic manufacture is not commercially feasible or where reasonable but unsuccessful efforts were made to find a domestic licensee.

This clause has real teeth in licensing negotiations and is often the provision a foreign licensee or a company with offshore manufacturing discovers late.

Confidentiality. 35 U.S.C. § 205 permits agencies to withhold invention disclosures from public disclosure for a reasonable time to allow patent filing — an important protection against premature publication through public records requests.

Stanford v. Roche and the assignment problem

For nearly thirty years, many institutions assumed Bayh-Dole vested title in the contractor automatically. Board of Trustees of the Leland Stanford Junior University v. Roche Molecular Systems, Inc., 563 U.S. 776 (2011) held otherwise.

The facts are a cautionary tale. A Stanford researcher signed an institutional agreement promising to assign future inventions — "I agree to assign." He then visited a company and signed an agreement stating that he "will assign and does hereby assign" his rights in inventions made there. He returned to Stanford and made an invention using both federal funding and knowledge from the visit.

The Court held that the company's present-tense assignment took effect immediately, while Stanford's promise to assign created only an equitable obligation. Chief Justice Roberts wrote:

"Since 1790, the patent law has operated on the premise that rights in an invention belong to the inventor. . . . Although much in intellectual property law has changed in the 220 years since the first Patent Act, the basic idea that inventors have the right to patent their inventions has not."

And on the statute:

"Bayh-Dole does not confer title to federally funded inventions on contractors or authorize contractors to unilaterally take title to those inventions; it simply assures contractors that they may keep title to whatever it is they already have."

The fix, adopted universally: institutional agreements now use present assignment language — "I hereby assign" — executed at hiring and often reconfirmed for specific projects. Any institution or company still using "agree to assign" in its inventor agreements has a title problem waiting to be discovered in diligence.

The broader lesson extends well beyond universities. Employment agreements, contractor agreements, and consulting agreements should all use present assignment language, and should be reviewed for it.

How a university license actually happens

The path from laboratory to license has a recognizable shape.

1. Invention disclosure. The researcher submits a disclosure to the technology transfer office describing the invention, its stage of development, funding sources, publications and presentations planned or made, and inventors. The funding-source question drives the Bayh-Dole analysis; the publication question drives the patent bar analysis.

2. Assessment. The office evaluates patentability, market potential, and development stage. Most disclosures do not result in a patent filing; institutional resources are finite and a patent family costs a great deal over its life.

3. Protection. A provisional application is common, buying twelve months. Note the interaction with publication: the United States allows a one-year grace period under 35 U.S.C. § 102(b) for the inventor's own disclosures, but most other jurisdictions have absolute novelty requirements, so a presentation before filing forfeits foreign rights entirely. This is the single most common way university inventions lose value.

4. Marketing. The office identifies potential licensees, often starting with companies the inventor already knows. Non-confidential summaries are circulated; confidential disclosure follows an agreement.

5. Negotiation. Term sheet, then license. The structure depends on whether the licensee is an established company or a startup the inventor is founding.

6. Management. Milestone tracking, royalty reporting, patent cost reimbursement, and — where the licensee is not performing — the difficult conversation about termination.

Frequently asked questions

Do inventors share in royalties? Yes, by institutional policy rather than by statute — though 35 U.S.C. § 202(c)(7) requires nonprofits to share royalties with inventors. Institutional shares vary widely, commonly between a quarter and half of net income to the inventors, with the balance divided among the laboratory, the department, and the institution.

Can an institution assign a subject invention outright? Nonprofits generally may not assign without agency approval, except to an organization whose primary function is invention management. This restriction surprises companies proposing to buy a patent rather than license it.

What if the inventor leaves and takes the technology to a company? The institution's rights depend on its inventor agreement and on when conception occurred. This is precisely the Stanford v. Roche problem, and it is why present assignment language matters. Where the invention was conceived at the institution, the institution owns it regardless of where the inventor now works.

Does Bayh-Dole apply to large companies? The statute's default applies to nonprofits and small businesses. Large businesses receiving federal funding are typically subject to agency-specific clauses that often reach similar results, but the analysis runs through the specific funding agreement rather than through § 202.

Does the government own our invention? No, if the institution elects title properly and meets the deadlines. The government retains a paid-up worldwide license for government purposes, plus march-in and domestic manufacturing conditions.

Can we grant a truly exclusive license? Exclusive as against everyone except the United States government. Disclose this to licensees; it is a recurring source of surprise.

What is a statutory bar? Under 35 U.S.C. § 102, certain disclosures start a clock or destroy novelty. The United States grace period does not exist in most other countries, so publication before filing generally forfeits foreign rights.

Are software and data subject inventions? Only if patentable. Copyrightable software and unpatented data are governed by institutional policy and by the terms of the funding agreement, and the rules differ substantially by agency.

What happens if we miss the disclosure deadline? The agency may take title. In practice extensions are frequently granted, but the risk is real and repeated lateness is a compliance problem.

Who pays for the patents? Initially the institution; in a license, the licensee typically reimburses past costs and pays ongoing prosecution costs. This is a negotiated term and a meaningful cost for a startup.

The license: what gets negotiated

University license agreements have a recognizable structure and a set of terms that reliably consume negotiating time.

Scope

Field of use. Institutions license narrowly and by field so that a single invention can support multiple licensees in non-overlapping markets. Companies want breadth. The resolution is usually a defined field with an option or right of first negotiation to expand on defined terms.

Territory. Worldwide is common; regional licensing occurs where different licensees serve different markets.

Exclusivity. Exclusive, co-exclusive, or non-exclusive. Exclusivity is the currency universities trade for development commitments, and the two should be negotiated together: a licensee that wants exclusivity across all fields should expect milestones across all fields.

Sublicensing. Permitted or not; if permitted, what share of sublicense income flows to the institution. Sublicense income sharing is one of the more contested economic terms, and the percentage typically declines as the licensee adds its own value.

Reserved rights. Institutions reserve the right to practice the invention for research and educational purposes, and to permit other nonprofit institutions to do so. Companies sometimes resist; institutions almost never yield, because the alternative is licensing away the ability to do further research on their own invention.

Economics

"Net sales" is where the money is. Deductions for returns, discounts, freight, insurance, and taxes are standard. Deductions for affiliate transfer pricing, bundled products, and combination products are where the disputes happen. Define combination product allocation explicitly rather than leaving it to a later argument.

Diligence and milestones

This is the heart of the institutional interest, and it is not merely commercial. Under 35 U.S.C. § 203, failure to take effective steps toward practical application is itself a march-in trigger, so an institution that licenses exclusively and then permits the licensee to sit on the technology has a compliance exposure as well as a lost opportunity.

Workable milestone provisions:

• Specific, dated, and objectively verifiable — "complete a Phase I trial by [date]," not "use commercially reasonable efforts"
• Tied to a development plan attached as an exhibit and updated annually
• With a cure mechanism and a defined consequence: conversion to non-exclusive, field reduction, or termination
• With an amendment process for genuine scientific delays, because research does not follow schedules

"Commercially reasonable efforts" standing alone is a term both sides regret. It is unmeasurable, it produces disputes, and it does not satisfy the institution's Bayh-Dole obligations.

Bayh-Dole flow-through

Every license of a subject invention must carry the statutory conditions forward:

• Acknowledgment of the government's paid-up license
• The domestic manufacture commitment under 35 U.S.C. § 204, or the licensee's obligation to cooperate in seeking a waiver
• Cooperation with reporting obligations, since the institution must report utilization it can only learn from the licensee
• Acknowledgment of march-in rights
• Requirement that the government support statement appear in patent filings

Licensees negotiating these should not expect movement. Institutions cannot waive statutory conditions.

Other recurring terms

Indemnification and insurance. Institutions require broad indemnity and named-insured coverage. This is close to non-negotiable and is often the first surprise for a startup licensee.

Patent prosecution control. Who instructs counsel, who selects countries, who decides about continuations. Licensees paying the costs want control; institutions want to preserve claims covering other fields. The usual compromise gives the licensee input and consultation rights with institutional control, or licensee control with institutional approval over abandonment.

Enforcement. Who may sue infringers, who controls, how recoveries are shared, and what happens if the licensee declines to enforce. Standing is a genuine issue: an exclusive licensee with all substantial rights may sue alone, while a licensee with a narrower grant must join the institution.

Termination. For breach with cure, for insolvency, and — commonly — at the licensee's convenience on notice. Address what happens to sublicensees on termination; a sublicensee that loses its rights because the head licensee breached has a serious problem, and a survival provision converting sublicenses to direct licenses is the standard fix.

Sponsored research and the ownership fight

When a company funds research at an institution, the parties are negotiating about inventions that do not yet exist. This produces the most predictable set of disagreements in the field.

The company wants ownership of what it paid for, or at least an exclusive license on defined terms, with the right to review publications and delay them.

The institution needs to retain ownership of inventions made by its personnel, to preserve its ability to publish, to avoid jeopardizing its tax status, and — if federal funds are involved — to comply with Bayh-Dole.

Where the compromise usually lands:

• Institution owns inventions made by its personnel; company owns inventions made by its personnel; joint inventions are jointly owned, with a mechanism for exploitation
• Company receives an option to negotiate an exclusive license, exercisable within a defined period after disclosure, on commercially reasonable terms to be negotiated in good faith — or, better for the company, on pre-agreed terms set out in the research agreement
• Publication review of 30 to 90 days to allow patent filing and removal of the company's confidential information; no right to suppress results
• Background intellectual property stays with its owner, with a license to the extent necessary to practice the funded results

The tax dimension matters and is often overlooked. Exempt organizations under 26 U.S.C. § 501 must ensure that sponsored research arrangements do not confer impermissible private benefit and do not generate unrelated business taxable income under 26 U.S.C. § 511. Research conducted for a sponsor's exclusive commercial benefit, on terms not available to others, raises both issues. Institutions therefore resist company ownership of results not because of stubbornness but because the arrangement can threaten exempt status and, where tax-exempt bond financing built the facility, can create private business use problems for the bonds.

Two structures to avoid. A "work for hire" characterization of research results — inventions are not works of authorship, and the label misleads. And an agreement that gives the sponsor rights in all inventions made in the laboratory during the term, which sweeps in unfunded and separately funded work and is unadministrable.

Startups from university research

A large share of institutional licensing goes to companies formed to commercialize the technology, often with the inventor as founder. This raises issues an arm's-length license does not.

Conflict of interest. The inventor is simultaneously a faculty member, an inventor with a royalty interest, an equity holder, and often an officer of the licensee. Institutions manage this through conflict-of-interest committees, disclosure requirements, restrictions on supervising students on sponsored work funded by the company, and sometimes limits on the founder's role in negotiating the license.

Equity. Institutions commonly take equity, typically a low single-digit to low double-digit percentage at formation, in lieu of higher cash consideration. Terms to address: anti-dilution (usually none beyond a defined financing threshold), conversion on a change of control, information rights, and whether the institution takes board representation (usually not, to avoid fiduciary complications).

Patent costs. A startup with no revenue cannot reimburse a $400,000 international patent family. Deferral, capping annual reimbursement, or converting arrears to equity are common accommodations.

Milestones calibrated to reality. A milestone schedule modeled on an established pharmaceutical company will be missed by a company with eighteen months of runway. Tie milestones to financing events as well as scientific ones.

Student and postdoctoral involvement. Work performed by students raises questions about their inventorship, their assignment obligations, and whether their thesis work is being directed by commercial rather than academic considerations. Institutions increasingly require explicit disclosure and management plans.

Use of institutional facilities. A startup operating in university space under a subsidized arrangement raises private benefit questions again, and — where the building was financed with tax-exempt bonds — private business use limits that are unforgiving.

The exit. On acquisition, the acquirer will diligence the license. Assignment provisions, change-of-control consents, milestone compliance, and Bayh-Dole compliance all get examined, and problems found then are expensive. Institutions that maintain clean compliance records add real value to their licensees at exit.

A worked case

Dr. Yelena Marchetti-Oduya runs a materials laboratory at a public research university. Her group develops a nanoporous membrane that separates lithium from brine at roughly half the energy cost of existing methods.

The funding is mixed: a five-year federal grant, a two-year foundation award, and $180,000 of industry funding from Crestmoor Resources, a mining company, under a sponsored research agreement signed eighteen months ago.

The disclosure

Yelena discloses to the technology transfer office on 3 March. The two-month clock under 37 C.F.R. Part 401 starts that day, not when the membrane worked.

The licensing officer, Ambrose Nakagawa, asks the four questions that determine everything:

"When was it conceived?" Yelena's notebooks show the key insight recorded on 11 January. This matters because Crestmoor's agreement covers inventions conceived during its funding period, which began the previous May.

"Which funds supported the work?" All three. The federal grant funded the graduate student who ran the synthesis; the foundation funded the characterization equipment; Crestmoor funded the brine-specific testing.

"Has any of this been published or presented?" Yelena gave a poster at a conference on 2 February and has a manuscript under review.

"Who are the inventors?" Yelena, a postdoctoral researcher, and — Ambrose asks specifically — anyone at Crestmoor who contributed to conception. Two Crestmoor engineers participated in the brine testing design. Whether they contributed to conception of the claimed invention is a legal question Ambrose flags for patent counsel, because joint inventorship with a company changes the ownership picture entirely.

The problems, in order of urgency

Foreign rights are already at risk. The 2 February poster was a public disclosure. The United States grace period under 35 U.S.C. § 102(b) preserves domestic rights for a year, but most other jurisdictions apply absolute novelty. Foreign rights are likely already lost unless a filing predated the poster, which it did not. Ambrose files a provisional immediately anyway, and documents the loss so nobody discovers it during a license negotiation two years later.

This is the single most common way university inventions lose most of their value, and it happens because researchers do not know that a poster is a publication.

Bayh-Dole deadlines. Federal disclosure by 3 May. Election within two years. Ambrose calendars both and files the agency disclosure in April.

Crestmoor's rights. The sponsored research agreement gives Crestmoor an option to negotiate an exclusive license in the field of "mineral extraction," exercisable within 90 days of disclosure. Ambrose notifies Crestmoor on 10 March.

Inventorship. Patent counsel interviews the Crestmoor engineers. Their contribution was execution of a testing protocol Yelena designed, not conception. They are not inventors. Ambrose documents the analysis in writing, because an inventorship error is a validity problem that surfaces in litigation.

The negotiation

Crestmoor exercises its option and wants: exclusivity worldwide, all fields, perpetual, with no milestones and a 1% royalty.

Ambrose's counter, and the reasoning behind each element:

Field limited to mineral extraction from brines. Crestmoor's option covers "mineral extraction." The membrane may also work for water purification and battery recycling, and those fields belong to other licensees. Crestmoor gets a right of first negotiation on adjacent fields, not a grant.

Exclusive, subject to the government license. Ambrose discloses expressly that the exclusivity is subject to the United States government's paid-up worldwide license under 35 U.S.C. § 202(c)(4), and to the reserved research rights of the university and other nonprofits. Crestmoor's counsel had not focused on this.

Domestic manufacturing. Crestmoor plans to manufacture membranes in a facility abroad. Under 35 U.S.C. § 204, no exclusive right to use or sell in the United States may be granted unless products are manufactured substantially domestically. Crestmoor must either commit to domestic manufacture or apply for a waiver, and the waiver requires showing that domestic manufacture is not commercially feasible. This term nearly ends the negotiation, and it is the provision foreign-manufacturing licensees discover late with dismaying regularity.

Milestones, dated and objective. Pilot-scale demonstration at 100 liters per hour within 24 months; a commercial installation within 60 months; first commercial sale within 84 months. Failure converts the license to non-exclusive after a 180-day cure period.

Economics. $75,000 upfront; escalating annual maintenance creditable against royalties; 3.5% on net sales with a defined combination-product allocation; milestone payments at pilot and commercial installation; patent cost reimbursement of past costs at signing and ongoing costs as incurred.

Reporting flow-through. Crestmoor must supply the utilization information the university needs to satisfy its own reporting obligations. This is not optional and Ambrose says so.

Where it lands

Crestmoor accepts the field limitation and the milestones, negotiates the royalty to 3%, and — after four months — commits to manufacturing membranes at a facility in Nevada rather than seeking a § 204 waiver, having concluded that the waiver application was slower and less certain than building domestically.

The government's march-in rights under 35 U.S.C. § 203 are acknowledged in the agreement, and Ambrose explains what they mean: no agency has ever exercised them, but the milestone structure exists partly so the university can demonstrate that effective steps toward practical application are being taken.

Foreign rights remain lost. Crestmoor's diligence flags it, and it reduces the upfront payment by roughly a third. Yelena now gives a fifteen-minute talk to incoming graduate students every autumn about what counts as a publication.

Material transfer and other everyday agreements

Most institutional agreements are not licenses. They are the small documents that move materials and information, and they create rights and obligations that surface years later.

Material transfer agreements. Governing the transfer of biological materials, compounds, cell lines, and devices between institutions or to companies. The terms that matter: whether the recipient may use the material only for the stated research; whether modifications and derivatives belong to the provider; whether the provider gets rights in inventions made using the material (a "reach-through" claim that institutions resist and companies request); publication rights; and whether the material may be transferred onward.

Reach-through provisions are the fight. A provider asking for ownership of, or a royalty on, anything invented using its material is asking for compensation disproportionate to a research reagent. Institutions generally refuse ownership and sometimes accept a narrow option to negotiate. Companies providing proprietary compounds have a stronger case and frequently get more.

Confidential disclosure agreements. Short, frequent, and consequential. Watch: the definition of confidential information (must exclude independently developed and publicly known information); the term of the obligation; whether the agreement inadvertently restricts the researcher's ability to publish or to work in the field; and residuals clauses, which institutions generally will not accept because they permit use of anything remembered.

Data use agreements. Increasingly important and often mishandled. Terms governing human subjects data, de-identification standards, permitted analyses, publication, and — now routinely contested — whether the data may be used to train machine learning models. That last question should be answered expressly rather than inferred from a general research-use clause.

Equipment loans and facility use. Where a company places equipment in a university laboratory, address who owns improvements, whether the company gets rights in results, and — critically — whether the arrangement creates private business use in a bond-financed facility.

Visiting scientist agreements. The Stanford v. Roche fact pattern. Anyone entering the laboratory should sign an agreement with present assignment language, and anyone leaving to visit another institution or company should be reminded that what they sign there may affect what they own here.

Federal laboratories and CRADAs

Bayh-Dole governs contractors. Federal laboratories operate under a parallel framework that companies encounter with some frequency and understand less well.

Cooperative Research and Development Agreements, authorized by 15 U.S.C. § 3710a, let a federal laboratory and a private party collaborate. The government contributes personnel, facilities, and equipment but generally not funds; the company may contribute funds. Key features:

• The collaborating party may obtain an exclusive or partially exclusive license in inventions made by laboratory employees under the agreement, negotiated in advance.
• The government retains a paid-up license for government purposes.
• Domestic manufacturing commitments apply, as under § 204.
• The laboratory may protect information produced under the agreement from disclosure for a period, which is the practical answer to the concern that everything given to a federal laboratory becomes public.

Licensing federally owned inventions — those where the government holds title — proceeds under 35 U.S.C. § 209, which permits exclusive licenses subject to findings that exclusivity is a reasonable incentive, that the scope is no greater than necessary, and that the applicant has a plan for development. Public notice and an opportunity for objection are typically required, and 35 U.S.C. § 207 authorizes agencies to obtain and license patents on federally owned inventions.

Practical differences from a university license: the negotiation is with a government entity subject to statutory findings rather than with an institution exercising commercial judgment; the timeline is longer; publication of the licensing opportunity may attract competing applicants; and the terms are constrained by regulation in ways a university's are not.

The uniform clauses. 35 U.S.C. § 206 directs the promulgation of standard patent rights clauses, and the resulting provisions in 37 C.F.R. Part 401 are what actually appear in funding agreements. Read the clause in the specific award; agencies have supplemental terms, and the clause — not the statute — is what binds the recipient.

Compliance in practice

Bayh-Dole compliance is administrative, unglamorous, and audited.

The reporting burden. Institutions report invention disclosures, elections, filings, patent issuances, and utilization through an electronic system. A large research university files thousands of transactions annually. Missed reports are compliance findings; patterns of them affect the institution's relationship with agencies and can surface in audits under 2 C.F.R. Part 200.

The government support statement. Every patent application on a subject invention must include a statement identifying the federal support and the agency. Omission is a defect that can be corrected, but a portfolio with systematically missing statements is a diligence finding and, in litigation, an invitation to argue the government has rights nobody disclosed.

Utilization reporting requires licensee cooperation. The institution cannot report what it does not know. Every license must obligate the licensee to supply utilization information, and technology transfer offices must actually collect it — a task that competes badly for attention against new deals.

Title and chain of title. Present assignment language in inventor agreements; confirmatory assignments recorded; joint inventorship resolved and documented; and a clear record for any invention where the funding analysis was close. Chain-of-title problems are the most common reason a university patent turns out to be unenforceable, and they are discovered at the worst possible time.

Waiver applications. Where domestic manufacture is not feasible, the waiver process under 35 U.S.C. § 204 requires a showing of reasonable but unsuccessful efforts to find a domestic licensee, or that domestic manufacture is not commercially feasible. Build the record contemporaneously — a waiver application supported by evidence assembled after the fact is weaker and slower.

Audit readiness. The documents an auditor will ask for: the disclosure log with dates; election records; filing records with government support statements; reporting confirmations; license agreements with Bayh-Dole flow-through provisions; utilization reports received from licensees; and the conflict-of-interest management plans for faculty-founded licensees. An institution that can produce these on request has a program; one that must reconstruct them does not.

Related documents

• Licensing University Technology: A Practical Guide for Companies and Inventors
• Bayh-Dole Compliance Checklist: A Practical Checklist
• Technology Transfer Toolkit: Invention Reporting, License Terms, and Sponsored Research
• How to License Your Patent: From Valuation to Term Sheet
• Patent Ownership, Assignments, and Standing: Chain of Title Problems That Sink Cases
• Employee Invention Assignment Agreements: Drafting for Enforceability Across Jurisdictions
• Nonprofit Governance, Unrelated Business Income, and Private Inurement
• Startup Formation and Fundraising Toolkit: A Roadmap and Resource Guide
• IP Transactions and Agreements Toolkit