Nearly every university patent license agreement contains an indemnification clause, and nearly every one of them runs in only one direction: the licensee agrees to indemnify, defend, and hold harmless the university against third-party claims, while the university refuses to offer any comparable protection in return. This asymmetry surprises many first-time licensees, particularly startup founders negotiating their first license from a university technology transfer office (TTO), who expect the give-and-take of a typical commercial contract. This guide explains what an indemnification clause does, why universities consistently decline to indemnify licensees, what universities offer instead, and where a licensee actually has room to negotiate. For the full clause-by-clause structure of a license agreement, see CASRAI’s guide to License Agreement Structure: Grant of Rights, Royalties, and Key Clauses, which covers indemnification as one of twelve standard clause categories; this guide goes deeper into that single clause.
What an Indemnification Clause Does
An indemnification clause allocates the risk and cost of third-party claims between the contracting parties. In a license agreement, the clause typically obligates one party to (1) defend the other against a covered claim, (2) pay any resulting settlement or judgment, and (3) reimburse the other party’s legal costs in connection with that claim. It is distinct from a warranty, which is a promise about the state of the licensed technology itself (for example, that the university has the right to license the patent); an indemnification clause instead addresses who pays when someone outside the contract — a customer, a patient, a competitor alleging infringement — sues over something arising from the licensed technology’s use, manufacture, or sale.
The Asymmetry: Licensee-to-University, Not University-to-Licensee
In the large majority of university patent license agreements, the indemnification obligation runs exclusively from the licensee to the university. The licensee agrees to indemnify the university (and typically its trustees, officers, employees, and inventors) against claims arising out of the licensee’s exercise of the license — product liability suits, claims that a licensed product injured someone, employment claims tied to the licensee’s own manufacturing operations, and similar third-party actions. The university, in turn, typically indemnifies nothing. This pattern is documented consistently across publicly available university license templates and legal-practice guidance for startups licensing from universities, including commentary from law firms that regularly represent university spinout companies (for example, Buchanan Ingersoll & Rooney’s guidance for startups licensing university technology) and template license agreements published by individual technology transfer offices, including those at Harvard, the University of Michigan, Mississippi State University, and the Johns Hopkins University Applied Physics Laboratory. The specific clause language differs institution to institution, but the underlying allocation — licensee indemnifies university, university indemnifies no one — is close to universal in the U.S. academic tech-transfer sector.
Why Universities Refuse to Indemnify Licensees
The refusal is not an oversight or a negotiating opening position that TTOs expect to give up under pressure; it reflects several structural features of what a university is and how it operates.
The university does not control the product
A university licenses a patent, know-how, or research tool; it does not manufacture, market, or sell a finished product. Once a licensee takes the technology and builds a product around it, decisions about design, manufacturing process, quality control, labeling, marketing claims, and distribution belong entirely to the licensee. From the university’s perspective, indemnifying the licensee would mean accepting financial responsibility for risks created by decisions the university had no part in and no ability to oversee — the opposite of how liability is conventionally allocated, where the party in control of a risk bears the cost of that risk.
Sovereign immunity, for public institutions
Many major U.S. technology transfer offices sit inside public, state-affiliated universities, which can claim sovereign immunity under the Eleventh Amendment and analogous state-law doctrines. State institutions frequently include express non-waiver language stating that nothing in the license agreement waives the institution’s sovereign immunity. Agreeing to indemnify a private licensee — effectively guaranteeing to pay a private party’s future legal liabilities — runs against that posture and, for many public universities, against state law and appropriations rules that restrict a state entity’s ability to pledge state funds to cover a third party’s contingent future liabilities without specific legislative authorization.
No revenue stream to absorb the risk
A commercial licensor that manufactures and sells a product earns revenue from that product and can price product-liability risk into what it charges, or self-insure against it. A university’s return from a license is typically a modest running royalty, and it has no comparable revenue base tied to the licensed product’s ultimate commercial risk. Underwriting open-ended indemnification obligations across a TTO’s entire license portfolio, most of which involve early-stage, unproven technology, would expose the university’s general funds — tuition, endowment, state appropriations, research funding — to product liability it played no role in creating.
Institutional risk management and insurance structure
University risk management offices typically carry general liability, property, and (where applicable) clinical/research liability coverage sized around the university’s own operations — teaching, research, campus facilities — not around downstream commercial products built by third parties. Agreeing to indemnify licensees would require either purchasing a fundamentally different category of coverage across an entire licensing portfolio or self-insuring against an unbounded and unpriced set of future claims, neither of which fits how university risk management is structured or funded.
Nonprofit and charitable-purpose considerations
Most U.S. research universities operate as nonprofit or state charitable institutions. Boards and general counsel offices generally treat open-ended financial guarantees to private, for-profit licensees as inconsistent with fiduciary obligations to use institutional assets for the university’s educational and research mission, reinforcing the reluctance to accept indemnification exposure on behalf of a commercial partner.
What Universities Offer Instead
Rather than indemnifying, universities typically limit their own exposure through a set of standard disclaimers that appear alongside the licensee-to-university indemnification clause:
- “AS IS” licensing, with no warranty of merchantability or fitness for a particular purpose — the university licenses the technology in its existing state and does not promise it will work for the licensee’s intended commercial application.
- No warranty of non-infringement — the university generally does not represent that practicing the licensed patent will avoid infringing a third party’s rights; freedom-to-operate analysis is left to the licensee.
- A narrow representation of authority to license — most agreements do include a limited representation that the university owns or controls the licensed patent and has the right to grant the license, which is a much narrower promise than a warranty about the technology’s performance or freedom to operate.
- A disclaimer of consequential, incidental, and punitive damages — even where the university does accept some narrow liability (for example, for its own gross negligence or willful misconduct, which some agreements do carve back in), that liability is typically capped and excludes indirect damages.
This combination — broad licensee-to-university indemnification, no university-to-licensee indemnification, and a disclaimer-heavy warranty section — is the standard risk allocation described in CASRAI’s companion guide to license agreement structure, under its Indemnification and Insurance and Representations, Warranties, and Freedom to Operate sections.
What Licensees Are Required to Do Instead
In exchange for the university’s refusal to indemnify, license agreements place two obligations on the licensee:
- Indemnify and defend the university against third-party claims arising from the licensee’s (and any sublicensee’s) manufacture, use, marketing, or sale of licensed products, including product liability claims, and reimburse the university’s associated legal costs.
- Carry adequate insurance — commercial general liability coverage from the outset, and product liability coverage once the licensee begins selling a product, at minimum coverage levels the university specifies, naming the university (and often its trustees and inventors) as an additional insured. Licensees are typically required to provide a certificate of insurance and to maintain coverage for a period after the license terminates, since product liability claims can arise from sales that occurred years earlier.
Early-stage startups sometimes negotiate a phased insurance schedule — lower coverage limits before a product reaches market, stepping up once sales begin — since carrying full product liability coverage before there is a product to sell is often impractical and unavailable at reasonable cost. This is one of the few genuinely negotiable elements of the insurance and indemnification package; the underlying obligation to indemnify the university is rarely negotiable at all.
Where Licensees Actually Have Room to Negotiate
Given that a university’s refusal to indemnify is close to a fixed position, licensee counsel typically focus negotiating effort elsewhere within the same clause family rather than on reversing the university’s stance:
- Scope of the indemnification obligation — narrowing it to claims “arising out of” the licensee’s own use, manufacture, or sale, rather than any claim merely “related to” the license, and excluding claims caused by the university’s own gross negligence or willful misconduct.
- Procedural protections — requiring prompt notice from the university of any claim, giving the licensee control of the defense (subject to the university’s approval of any settlement that admits liability or imposes obligations on the university), and a right to participate in choosing defense counsel.
- Insurance minimums and timing — negotiating coverage amounts appropriate to the technology’s actual risk profile and a realistic date by which product liability coverage must be in place, rather than accepting boilerplate figures drafted for a higher-risk technology class (medical devices and therapeutics typically warrant materially higher minimums than, say, licensed software or research tools).
- A narrow, factual representation regarding known claims — some universities will represent that, to their actual knowledge, no third party has asserted an infringement claim against the licensed patent as of the effective date. This is far short of a warranty of non-infringement, but it is a fact-based statement about the university’s own knowledge that many TTOs will accept, unlike an open-ended risk-shifting promise.
What licensees should not expect to negotiate away, in almost any circumstance, is the core allocation itself: a university indemnifying a licensee against product liability or infringement claims arising from the licensee’s own commercial activity. Licensees who encounter unusually favorable terms on this point should treat it as a signal to re-read the clause carefully rather than assume it reflects standard market practice.
How This Interacts With Other License Clauses
Indemnification does not stand alone in the agreement. It typically survives termination for claims tied to pre-termination sales (see the Term and Termination discussion in CASRAI’s License Agreement Structure guide), flows through to sublicensees — a licensee’s sublicense agreements generally must require the sublicensee to indemnify on terms no less protective than the head license — and interacts with field-of-use and exclusivity terms: an exclusive licensee taking on broader commercial rights across a defined field of use is typically expected to carry correspondingly broader insurance and indemnification obligations than a narrow, non-exclusive licensee. It is negotiated alongside, and is conceptually related to, the due diligence questionnaire a prospective licensee completes early in the process, since the university’s assessment of a licensee’s financial capacity to actually make good on an indemnification promise (and to obtain the required insurance) factors into how firmly the TTO holds its position on these terms.
Frequently Asked Questions
Will a university ever agree to indemnify a licensee?
It is uncommon. Some agreements carve back narrow university liability for the institution’s own gross negligence or willful misconduct, and a university may make a limited representation that it has the right to grant the license. A general commitment to indemnify the licensee against third-party claims, however, is rarely available, for the structural reasons described above — sovereign immunity for public institutions, lack of control over the downstream product, and the absence of a revenue base to underwrite the risk.
What is the difference between indemnification and a warranty of non-infringement?
A warranty of non-infringement would be a promise that practicing the licensed patent does not infringe a third party’s rights. Indemnification is a promise to pay if someone sues over that (or another) issue. Universities typically decline both, but they are legally distinct: a licensee can occasionally obtain a narrow, knowledge-qualified representation on infringement even where indemnification remains off the table entirely.
Can a public university legally indemnify a licensee even if it wanted to?
Often not without difficulty. Many state universities operate under state law and appropriations rules that restrict a state entity’s authority to pledge public funds to cover a third party’s future contingent liabilities, and public institutions frequently preserve sovereign immunity explicitly in the agreement. This is a separate and additional obstacle on top of the university’s own risk-management preference not to indemnify.
What insurance does a licensee typically need to carry?
Commercial general liability coverage from the license’s effective date, and product liability coverage once the licensee has a product on the market, at coverage minimums the university specifies (commonly in the low millions of dollars per occurrence for higher-risk technology, though figures vary by institution and field), with the university named as an additional insured and proof of coverage provided on request.
Does this pattern apply outside patent licenses — for example, software or biological material licenses?
The same basic asymmetry appears across most university IP licenses, not only patents, though the specific insurance minimums and claim categories differ. CASRAI’s dictionary entry on the Software License Agreement and guide to license agreement structure cover how indemnification provisions are adapted for non-patent IP types.
