University and Research Institution IP Toolkit: Sponsored Research, Bayh-Dole, and Spin-Outs

By Casey Scott McKay · 2023-10-10

A company negotiating with a university for the first time usually assumes it is negotiating with a company, and roughly two-thirds of the terms it objects to are statutory obligations rather than positions. This toolkit sorts them. It covers the Bayh-Dole framework and what it actually requires, why Stanford v. Roche made assignment language the decisive question, and the diligence that verifies title rather than accepting it. It works sponsored research options and the publication mechanism that functions, material transfer agreements and the reach-through terms to refuse, and the exclusive licence terms that determine whether a spin-out can finance. It closes with equity and anti-dilution, conflict management, federal laboratory agreements, inter-institutional arrangements, and an escalation approach.

IP and Technology > IP and IT in Corporate Transactions | Toolkit | Published 10 October 2023 - Updated 28 March 2026 | Casey Scott McKay - marksy.us

Summary. A company negotiating with a university for the first time usually assumes it is negotiating with a company, and roughly two-thirds of the terms it objects to are statutory obligations rather than positions. This toolkit sorts them. It covers the Bayh-Dole framework and what it actually requires, why Stanford v. Roche made assignment language the decisive question, and the diligence that verifies title rather than accepting it. It works sponsored research options and the publication mechanism that functions, material transfer agreements and the reach-through terms to refuse, and the exclusive licence terms that determine whether a spin-out can finance. It closes with equity and anti-dilution, conflict management, federal laboratory agreements, inter-institutional arrangements, and an escalation approach.

Keywords: technology transfer toolkit · Bayh-Dole compliance · subject inventions · election of title · march-in rights · US manufacturing preference · Stanford v Roche · sponsored research options · publication delay · material transfer agreements · reach-through refusal · exclusive licence terms · diligence milestones · sublicensing share · patent cost reimbursement · spin-out equity · anti-dilution · conflict management · CRADA · inter-institutional agreements

Start Here

A company negotiating with a university for the first time usually assumes it is negotiating with a company.

It is not. It is negotiating with an institution that cannot assign certain rights because a statute forbids it, that cannot accept publication restrictions because its faculty would revolt, that must share royalties with inventors by law, and whose negotiator's incentives are not commercial in any recognisable sense.

Understanding why the institution's positions are what they are converts most of these negotiations from adversarial to procedural. Roughly two-thirds of the terms a company finds objectionable are not negotiating positions at all — they are statutory obligations, and pushing on them wastes months.

The other third is genuinely negotiable, and knowing which third is which is the whole skill.

Sort every term into one of three buckets before opening the redline.

Statutory — do not negotiate. The government's non-exclusive licence in any subject invention under 35 U.S.C. § 202. March-in exposure under 35 U.S.C. § 203. The US manufacturing preference under 35 U.S.C. § 204. Inventor royalty sharing. Disclosure, election, and reporting obligations under 37 C.F.R. § 401.14. And, for public institutions, statutory limits on indemnification, governing law, and forum.

Institutional policy — negotiable with escalation. Publication mechanics. Research and educational reservations. Standard royalty and sublicensing shares. Equity policy in spin-outs. Conflict management. Warranty disclaimers.

Genuinely open — where to spend the effort. Field of use. Diligence milestones and their consequences. Sublicensing income share. Patent cost reimbursement and prosecution control. Improvements definition. Assignment on a change of control. Enforcement rights and recovery sharing. Territory carve-outs.

The question that saves weeks. When the institution says it cannot agree to something, ask: is that statute, state law, or policy? Offices answer honestly, and the answer tells you whether to escalate or move on.

This toolkit answers four questions. What does Bayh-Dole actually require? How is title verified rather than assumed? Which licence terms determine whether a company can finance? And what does a spin-out need beyond the licence?

The Bayh-Dole Framework

What it did. Before 1980, inventions made with federal funding generally belonged to the government and very few were commercialised. 35 U.S.C. § 200 and following allowed contractors — universities, small businesses, and nonprofits — to elect to retain title to inventions made with federal support.

A subject invention is one conceived or first actually reduced to practice in the performance of work under a funding agreement. 35 U.S.C. § 201 supplies the definitions and 35 U.S.C. § 202 sets out the right to elect title, subject to conditions.

The obligations are conditions rather than aspirations.

Disclosure to the funding agency within a set period after the inventor discloses internally. Late disclosure can forfeit title.

Election to retain title within a set period after disclosure. Failure to elect can forfeit title to the government, and there is no cure.

Filing within a set period after election, with notification requirements for foreign filing decisions.

The government licence. A non-exclusive, non-transferable, irrevocable, paid-up licence to practise the invention worldwide on the government's behalf. It runs with the patent and cannot be negotiated away.

March-in rights under 35 U.S.C. § 203, permitting an agency to require licensing where practical application is not achieved, where health or safety needs are unmet, where regulatory requirements are unmet, or where the manufacturing preference is breached. Never exercised despite periodic petitions, which does not make it irrelevant to a licensee's risk assessment.

The US manufacturing preference under 35 U.S.C. § 204, requiring products embodying a subject invention and sold in the United States to be manufactured substantially in the United States where an exclusive licence to use or sell is granted. Waivers exist and must be sought.

Royalty sharing with inventors, which is why institutional licences allocate revenue in ways that look unusual to commercial parties.

Utilisation reporting on commercialisation progress, the mechanism by which agencies monitor practical application.

The practical consequence for a licensee. The government licence and the manufacturing preference run with the patent and belong in the diligence file of any acquirer.

Why Stanford v. Roche Matters

Board of Trustees of Leland Stanford Junior University v. Roche Molecular Systems held that Bayh-Dole does not itself vest title in the contractor. It governs the allocation of rights between contractor and government, and presupposes that the contractor obtained title from the inventor in the ordinary way.

The facts. A researcher signed a university agreement promising to assign future inventions, then signed a company agreement stating that he "hereby assigns" rights in inventions made there. The present assignment took effect immediately; the promise did not. The company took title.

The lesson, in three words. Read the clause. Filmtec v. Allied Signal established the distinction.

What a licensee should verify. Not a representation that the institution owns the patent, but the inventors' assignments, their operative language, and their recordation under 35 U.S.C. § 261.

Where defects concentrate. Visiting researchers bound by a home institution's policy. Postdoctoral researchers on external fellowships. Students, whose obligations vary by institution and programme. Faculty with joint appointments, including at affiliated hospitals. And anyone who consulted for a company during the relevant period.

Where a defect exists, the cure requires the inventor's cooperation, and inventors who have left academia and understand their position negotiate accordingly.

Material Transfers

The governance point first. No scientist signs one. Route every material transfer agreement through counsel without exception, because a two-page document obtained to get a reagent can encumber a lead programme and surface four years later in diligence.

What to accept. A licence to use the material for the stated research purpose. A prohibition on onward transfer without consent. An acknowledgement obligation in publications. A non-exclusive licence back to the provider to use resulting research results for its own internal research and educational purposes. Return or destruction on completion.

What to refuse, and escalate rather than sign. Ownership of, or an option over, inventions the recipient makes using the material. Reach-through royalties on products developed using it. Publication approval rights. Any grant covering the recipient's background technology.

Why reach-through is objectionable in principle. The material may be a minor research input, and an obligation proportionate to a reagent is not one proportionate to a product.

The negotiating line that works. Offer a non-exclusive research licence to results and acknowledgement. That is proportionate, and most offices accept it when asked — the reach-through is usually in the form because nobody objected, not because the institution requires it.

Use standard forms. The uniform biological material transfer agreement and its simple letter variant are widely accepted, balanced, and recognised immediately, and proposing one is frequently faster than negotiating a bespoke form.

Chain of transfer. Where the provider obtained the material from a third institution, those onward transfer restrictions bind. Ask for the chain and confirm the provider had the right to transfer.

Human materials. Consent scope governs absolutely. Confirm the consent covers commercial research, because nothing downstream can expand it.

Track them. A register with material, provider, date, purpose, restrictions, and the programme it touched, reviewed before any transaction.

The Research Exemption Myth

What people believe. That academic research is exempt from infringement.

Madey v. Duke University holds the common law experimental use defence is very narrow, confined to acts performed for amusement, idle curiosity, or strictly philosophical inquiry. Research furthering an institution's legitimate business — attracting funding, students, and faculty — falls outside it.

Which means institutional research is not exempt, and institutions that assumed otherwise have been sued.

The statutory safe harbour. 35 U.S.C. § 271 at subsection (e)(1) exempts uses reasonably related to developing and submitting information under a federal law regulating drugs or biological products. Merck KGaA v. Integra Lifesciences read it broadly, covering preclinical work where there is a reasonable basis to believe the compound may be the subject of a submission.

Its limits. It covers regulatory submissions, not research tools used in the ordinary course, and the boundary is contested.

Practical consequence. An institution with a substantial research programme needs freedom-to-operate awareness like any other, and a company licensing from an institution should not assume the underlying research was cleared. Reagents, constructs, and instruments used in the work may themselves be licensed from third parties, and the licensee inherits nothing.

The Exclusive Licence

Accept quickly. The institutional research and educational reservation. Research use by other nonprofits. The government licence and march-in exposure where Bayh-Dole applies. The manufacturing preference. Warranty disclaimers covering validity, enforceability, non-infringement, and fitness. Limited or absent institutional indemnity, addressed through insurance requirements and a narrow mutual indemnity for each party's own negligence.

Negotiate hard on seven terms.

Field of use. Define by application, not by product, and address expressly what happens when the product evolves. A vague field produces disputes; a narrow one produces a competitor licensed in the adjacent field. Ask what other fields the institution intends to license and to whom.

Diligence milestones. Accept their existence — the institution's statutory objective is practical application, and unexercised exclusive licences are the failure mode its office is measured on. Negotiate three things: achievability against the company's real timeline, a meaningful cure period, and conversion to non-exclusive rather than termination as the consequence.

Why conversion rather than termination. A company that has invested years and capital should not lose everything for a slipped date, and the institution's real objective is served by conversion.

Sublicensing. Confirm the right exists without prior approval, or with approval not unreasonably withheld and deemed given after a period. Confirm sublicences survive termination where the sublicensee is not in breach — without that, a sublicensee's diligence finds its rights depend on the licensee's compliance, which is not financeable. Negotiate the income share, frequently set far above the running royalty.

Patent costs. Past costs are the largest early cash item. Negotiate a cap, an instalment schedule, consultation rights on prosecution, and the right to decline nominated jurisdictions with rights reverting there.

Improvements. The institution cannot commit future inventions by faculty generally. Define narrowly — inventions by the named inventors, within a defined period, dominated by or dominating the licensed claims — and accept that anything wider will be refused.

Enforcement. The licensee needs the right to sue, the institution's agreement to join as a necessary party at the licensee's expense, control including settlement authority, and a recovery-sharing formula. Address what happens if the institution declines to join.

Assignment. Change of control without consent, or with consent not unreasonably withheld. Investors require this and its absence is a financing problem rather than a legal nicety.

Insolvency. Confirm the licence survives and that Mission Product Holdings v. Tempnology protection or an equivalent contractual mechanism applies.

Valuing What Is Licensed

Do not count patents. Count claims that survive scrutiny.

Eligibility. Institutional portfolios cluster in life sciences and software, the two areas where 35 U.S.C. § 101 has narrowed most. Association for Molecular Pathology v. Myriad Genetics removed isolated natural DNA. Mayo Collaborative Services v. Prometheus Laboratories narrowed diagnostic methods. Alice v. CLS Bank International supplies the framework generally.

Read the file history. Eligibility rejections and the amendments made to overcome them tell you what the claims actually cover, and they are public.

Assess the disclosure. A provisional filed under conference-deadline pressure may not support the claims eventually pursued, and the priority date is only as good as the disclosure behind it.

Assess enablement. Institutional patents frequently claim broadly on limited working examples, which is where Amgen v. Sanofi and the written description cases bite.

Assess foreign coverage honestly. Where a publication preceded filing, foreign rights are usually gone regardless of what the licence says. Ask what was published, presented, posted, or deposited, with dates — this single question determines whether a worldwide exclusive licence is worth what it appears to be. Confirm the publication date under 35 U.S.C. § 122 and whether a non-publication request was made.

Ask what the technology consists of beyond the patents. Software already published to a public repository, where the release is generally irrevocable as to the code released. Datasets, with provenance, consent scope, and any data use agreement restricting transfer. Biological materials with their own transfer chains. And know-how — the unpublished practical knowledge held by a postdoctoral researcher, without which a patent licence is frequently a licence to something that does not work.

Negotiate consulting or transitional support explicitly, routed through the conflicts office rather than agreed privately.

The Spin-Out

Sequence the work. Conflicts first, licence second, equity third, financing fourth. Reversing that order produces a company whose licence must be renegotiated in front of investors.

Conflicts. Engage the institutional conflicts office before formation. Expect a management plan addressing the founder's roles, whether they may supervise students on company-relevant work, whether they may serve as an officer, how company-funded sponsored research to their own laboratory is reviewed, and what disclosures are required in publications.

Treat the plan as a deal term. A conflict handled badly damages the founder's career, the institutional relationship, and the company's access to the laboratory that produced the technology.

Equity. Institutions typically take a single-digit percentage. Negotiate anti-dilution hardest — protection through a Series A materially changes founder and early investor outcomes, and investors price it. Understand what the equity is: common or preferred, participating or not, voting or not, and whether a board seat or observer rights attach.

What investors check, in order.

Exclusive, worldwide, and in a field broad enough for the business plan.
Survives insolvency.
Assignable on a change of control without consent.
Sublicensing permitted, with sublicences surviving termination.
Diligence milestones achievable and currently met.
Patent costs quantified, capped, and scheduled.
Enforcement rights adequate with institutional cooperation committed.
Government rights and manufacturing preference disclosed.

Fix all eight before the term sheet. Each is fixable and institutions are generally cooperative, but amendments take a quarter and term sheets do not wait.

Facilities. A spin-out operating in institutional space needs express terms on inventions made there and by whom, or the ownership question the licence resolved reopens.

Students and postdoctoral researchers. Work on company problems raises both conflict and ownership questions, and their assignment obligations may differ from those of employees.

Federal Laboratories

Different statute, different vehicle. Government laboratories operate under the Federal Technology Transfer Act, and the instrument is the cooperative research and development agreement under 15 U.S.C. § 3710a.

What a CRADA gives. The laboratory contributes personnel, facilities, and equipment but generally not funds. The collaborator contributes funds and resources and receives an option to an exclusive licence in a defined field for inventions made under the agreement. Data produced may be protected from disclosure for a defined period, which is a significant benefit and a common reason for using the vehicle.

Licensing federally owned inventions is governed by 35 U.S.C. § 209, which requires a development plan and, for exclusive licences, public notice and an opportunity for objection.

Practical differences. Timelines are longer, terms more standardised, the negotiator has less discretion, and approval chains are longer. Budget accordingly and do not read delay as a negotiating signal.

The manufacturing preference applies, with a separate waiver process.

Export control applies to collaborations involving controlled technology or foreign nationals, independently of intellectual property terms, and laboratories have compliance offices that will raise it.

Inter-Institutional Arrangements

Multi-institution collaborations produce joint inventorship across institutions, which produces joint ownership — and under 35 U.S.C. § 262 each owner may license independently without accounting to the other.

Which is a result neither institution intended and which makes exclusivity illusory for a licensee.

Inter-institutional agreements exist to fix it. They designate one institution to lead prosecution and licensing, allocate revenue, and bind both.

Confirm one exists before licensing jointly owned technology. Without it, a licensee may need a licence from each owner, at combined terms neither would have obtained alone, and either owner may license a competitor.

Where none exists, make its execution a condition of the licence rather than a hope.

Inventorship accuracy matters more here than anywhere. Academic authorship conventions and patent inventorship are different systems — a paper's author list is not the inventor list, and the Pannu v. Iolab standard controls. Correction runs through 35 U.S.C. § 256, and an omitted unassigned inventor can license the patent to anyone.

Diligence on Institutional Technology

Eight questions, in order.

Does the institution hold title? Read the assignments and check recordation.
Were Bayh-Dole obligations met? Timely disclosure, election, filing, and utilisation reports, obtained in writing.
What runs with the patent? Government licence, march-in exposure, manufacturing preference, research reservations.
Is the licence assignable on a change of control?
Are diligence milestones current? A licence in breach is one the institution can convert or terminate, and licensees fall behind without notifying anyone.
What sponsored research or material transfer agreements touch the technology? Reach-through obligations and sponsor options surface here.
Who else has rights? Other licensees in other fields, nonprofit research licences, and prior options granted.
Are the inventors still cooperative? Correction, foreign formalities, and litigation support all require them.

The recurring finding. Missed utilisation reports and lapsed milestone compliance — both curable, both requiring the institution's agreement, and both far easier to obtain before a transaction is announced than during it.

Money and Who Gets It

Revenue allocation. Institutional licence income is typically split among the inventors, their departments, and central funds, under a policy predating any particular deal.

Inventor share is statutory in part. Bayh-Dole requires sharing with inventors for subject inventions, and institutional policy generally extends it to all inventions.

Why a licensee should care. The individual scientist has a direct financial interest in the licence's success — usually helpful, occasionally distorting, since an inventor may push for terms the office would not or resist a restructuring that reduces their share.

Sublicensing income. Frequently taken at a much higher rate than the running royalty, and a business model built on sublicensing lives or dies on this number.

Equity versus royalty. Institutions increasingly take equity in spin-outs in place of upfront cash. Understand what the equity is, because those terms drive institutional behaviour in later financings.

Patent cost reimbursement. Frequently the largest early cash obligation, covering costs incurred before the licensee existed.

Minimum annual royalties exist to ensure the licence produces something before commercialisation, and are negotiable in amount and in when they begin.

The economic reality. Very few institutional licences generate significant revenue, and offices know it. A posture that acknowledges this — modest upfront, meaningful milestones, fair royalty — closes faster than one fighting every number.

Escalation

Who handles what. A licensing associate handles standard terms with limited financial authority. A director or office head handles departures from policy. General counsel handles indemnity, liability, governing law, forum, and institutional risk. The vice president for research handles publication, conflicts, and faculty relations.

The principal investigator is not a negotiator but is an ally, and their support moves internal decisions counsel cannot.

How to escalate well. Present the commercial reason, not the legal position. "This term prevents us financing the company, and here is why an investor reads it that way" moves an office far better than a redline with no explanation.

What not to do. Escalate over a statutory term. It signals that nobody read the framework and costs credibility on the terms that are actually open.

Use comparables. Institutions benchmark against each other, and a term another institution granted in a similar deal is persuasive and will be checked.

Timing. Institutional decision cycles are slower than commercial ones — committees meet monthly, approvals route through several offices, and certain months are lost. Build the calendar into the transaction timetable rather than treating delay as resistance.

The relationship matters. Most companies that license from an institution once will do so again, and the office remembers how the last negotiation went.

For the Institution

The framework above reads from the company's side. The institution's problems differ.

Volume against budget. Far more disclosures arrive than can be filed, and filing decisions are made on thin information at the point where the technology is least developed. Provisional filings buy a year of information at modest cost, and a disciplined re-evaluation at the twelve-month mark is where most of the portfolio value sits.

Faculty relations. An office perceived as a gatekeeper receives fewer disclosures, and undisclosed inventions are lost outright. Speed of response matters more to that relationship than the eventual filing decision.

Compliance load. Disclosure, election, filing, and utilisation reporting run on statutory clocks across hundreds of assets and multiple agencies. A missed election under 35 U.S.C. § 202 is the failure with the worst consequence and the least visibility, and it is a docketing problem rather than a legal one.

Unexercised exclusives. The failure mode that matters institutionally is a technology exclusively licensed to a company doing nothing with it, which is why diligence milestones exist and why offices resist licensees who want them removed.

Measuring the office. Licence counts and revenue are the visible metrics and the misleading ones, since revenue is dominated by a handful of outliers. Disclosures received, response times, and startups formed describe the function better.

For a company negotiating with an office: the person opposite is measured on getting technology used, not on extracting maximum value from any single deal. A credible development plan is worth more in that negotiation than an aggressive financial position.

International Institutions

Do not port the US template. The ownership defaults, compensation obligations, and funding conditions differ enough that it will be wrong in ways that are not obvious.

Ownership. Several systems historically vested title in the individual academic rather than the institution. Most have moved away, but legacy inventions and some jurisdictions still follow it. Confirm who holds title.

Inventor remuneration. Many systems grant employee inventors statutory compensation surviving assignment, and academic inventors are covered.

Public funding conditions. National and regional programmes impose conditions on exploitation, manufacturing location, reporting, and sometimes where a resulting company is established — operating alongside Bayh-Dole where both apply.

Grace period. The US grace period under 35 U.S.C. § 102 has few equivalents. In most systems pre-filing publication is an absolute bar, which is why the publication delay is worth more abroad than at home.

Export control and research security. Collaborations involving controlled technology or foreign nationals raise licensing requirements independent of intellectual property terms.

Practical approach. Engage local counsel for the first agreement in any new jurisdiction and reuse what you learn. The second agreement costs a fraction of the first.

Three Deals

The sponsored research that produced nothing. A company funded a laboratory for three years under an agreement giving it an option to negotiate a licence to resulting inventions. Two patents issued. The option was exercisable within six months of disclosure, and both lapsed unexercised because nobody was tracking them — the scientist who managed the relationship had moved on. The institution subsequently licensed one of the patents to a competitor in the same field. Nothing was breached and nothing was recoverable. The fix costs nothing: diary the windows, name an owner, and pre-agree the licence framework so exercise is a decision rather than a negotiation.

The material transfer that encumbered a programme. A researcher at a biotechnology company obtained a mouse strain under a two-page agreement containing a reach-through royalty on any product developed using the material. The strain was used in one early experiment and abandoned. Four years later, during acquisition diligence, the agreement surfaced and the buyer required an opinion on whether the lead programme was captured. It was not, but proving it took months and a negotiated release the institution had no obligation to grant.

The spin-out that could not finance. A licence to a founder's new company was exclusive, worldwide, and adequately priced. It also prohibited assignment without consent, contained no change of control carve-out, permitted sublicensing only with approval, and set a first milestone eighteen months out that the company had already missed by the time it approached investors. Each defect was fixable and the institution was cooperative, but the amendments took a quarter and the term sheet was withdrawn. A university licence is a financing document and should be reviewed by whoever will present it to investors before it is signed.

Where Negotiations Stall

Publication. Resolved by treating it as notice plus a filing delay, not as approval.

Indemnity. Institutions cannot indemnify broadly and public institutions may be barred by statute. Resolve with insurance requirements and a narrow mutual indemnity for each party's own negligence.

Warranties. Institutions disclaim everything and will not change. Price the risk instead.

Ownership of improvements. The licensee wants them included; the institution cannot commit future inventions by faculty who may not be involved. Resolve with a defined category limited to the named inventors during a defined period.

Field of use. Genuine negotiation and worth the time.

Diligence milestones. Negotiate cure periods and conversion, not existence.

Assignment. Negotiate change of control without consent, which investors require.

Confidentiality. Institutions cannot commit their students and cannot restrict use of research results generally.

The general rule. Where the institution says it cannot do something, ask whether that is policy, statute, or state law. Policy is negotiable with escalation; statute and state law are not, and the answer saves weeks.

A Closing Note

The distinctive thing about institutional technology transfer is that many of the terms a commercial party finds strange are not commercial terms at all.

The government licence exists because a statute says so. The publication right exists because faculty independence is the institution's reason for being. The inventor royalty share exists because Bayh-Dole requires it. The refusal to indemnify may be state law. Arguing with any of these is a way of spending months and arriving where you started.

What is genuinely negotiable is narrower and more interesting: the field, the diligence consequence, the sublicensing share, the patent cost arrangement, improvements, assignment, and enforcement. Those seven determine whether the licence supports a business.

Verify title rather than accepting it, because Board of Trustees of Leland Stanford Junior University v. Roche Molecular Systems is the case it is for a reason. Ask what has been published and when, because the answer determines whether foreign rights exist. Ask what the technology consists of beyond the patents, because the know-how and the materials are frequently what makes it work.

And engage the conflicts office early in any spin-out, because the relationship with the laboratory is the asset the licence cannot replace.

A Suggested Reading Path

For the framework:

For the government contracting dimension:

For the ownership and inventorship questions:

For the patent strategy layer:

Primary Authorities

| Authority | Proposition | |---|---| | 35 U.S.C. § 200 | Bayh-Dole policy | | 35 U.S.C. § 201 | Subject invention definition | | 35 U.S.C. § 202 | Election of title; government licence | | 35 U.S.C. § 203 | March-in rights | | 35 U.S.C. § 204 | US manufacturing preference | | 35 U.S.C. § 205 | Confidentiality of disclosures | | 35 U.S.C. § 209 | Licensing federally owned inventions | | 35 U.S.C. § 261 | Assignment and recordation | | 35 U.S.C. § 262 | Joint owners | | 35 U.S.C. § 256 | Correction of inventorship | | 35 U.S.C. § 101 | Eligibility | | 35 U.S.C. § 102 | Novelty; grace period | | 35 U.S.C. § 122 | Publication at eighteen months | | 35 U.S.C. § 271 | Infringement; regulatory safe harbour | | 37 C.F.R. § 401.14 | Standard patent rights clause | | 15 U.S.C. § 3710a | Cooperative research and development agreements | | Board of Trustees of Leland Stanford Junior University v. Roche Molecular Systems | Bayh-Dole does not vest title | | Filmtec v. Allied Signal | Present assignment versus promise | | Madey v. Duke University | No general research exemption | | Merck KGaA v. Integra Lifesciences | Regulatory safe harbour scope | | Association for Molecular Pathology v. Myriad Genetics | Isolated DNA ineligible | | Mayo Collaborative Services v. Prometheus Laboratories | Diagnostic eligibility | | Alice v. CLS Bank International | Abstract idea framework | | Amgen v. Sanofi | Enablement of broad claims | | Pannu v. Iolab | Joint inventorship | | Mission Product Holdings v. Tempnology | Licence survives rejection |

Forms and Templates

The instrument that decides most institutional negotiations is not the licence but the term sheet annexed to the sponsored research agreement, because an option to negotiate is worth very little and an option to a licence on pre-agreed terms is worth a great deal. The License Agreement Template adapts to that purpose: licensed patents defined to include continuations and foreign counterparts; field defined by application rather than by product; exclusivity subject to the government licence and the institutional research reservation; improvements limited to the named inventors within a defined period; financial terms including a stacking provision and a sublicensing share negotiated separately from the running royalty; patent costs capped and scheduled with the right to decline nominated jurisdictions; diligence milestones with a cure period and conversion to non-exclusive rather than termination; sublicensing on notice with sublicences surviving termination; enforcement with joinder committed and recoveries shared; assignment permitted on a change of control; and survival on insolvency consistent with Mission Product Holdings v. Tempnology.

The Assignment Agreement Template is the diligence instrument. Verifying title means reading the inventors' assignments and classifying the operative language as present assignment, promise, or absent, then confirming recordation under 35 U.S.C. § 261 — and where a defect is found, the same template supplies the confirmatory assignment that cures it.

The Portfolio Inventory Template adapts to two registers this practice requires. The first is the option diary: every option held, its trigger, its expiry, and a named owner, because options lapsing unnoticed is the most common avoidable loss in the field. The second is the material transfer register: material, provider, date, purpose, restrictions, and the programme it touched — which is what allows a clean answer during diligence four years later.

Related Toolkits and Checklists

The Technology Transfer Checklist runs the relationship in phase order with gates before signature and before financing. The Patent Licensing and Technology Transfer Toolkit covers the ordinary licensing framework these institutional terms modify. The IP Audit and Portfolio Governance Toolkit covers the title verification method applied here to inventors' assignments. The Federal Contract IP Checklist covers the government contracting side of the same Bayh-Dole obligations. And the IP and Antitrust Toolkit covers the restraint questions that arise where an institution licenses the same technology into competing fields.