Why Most Co-Invention Disputes Destroy University-Industry Relationships Permanently

Co-invention disputes are one of the quietest deal-killers in university-industry research. They rarely make headlines. They don't show up in the statistics that tech transfer offices report to their boards. But ask anyone who has managed a sponsored research portfolio for more than a few years, and they'll have at least one story: a relationship that was fine until inventorship came up, and then was never quite fine again.

Photo by Pixabay on Pexels.

The mechanics are predictable. A company funds research at a university. Researchers on both sides work together closely, sharing data, brainstorming, running experiments. Something patentable emerges. Then the question arrives: who invented it?

That question sounds simple. It isn't.

Under U.S. patent law, inventorship is a legal determination tied to contribution to the conception of the claimed invention. Not who funded it. Not who ran the experiments. Not who had the idea first in a loose, conversational sense. The legal test is specific, and it's applied claim by claim. Two people can both contribute meaningfully to a project and yet only one qualifies as an inventor on a particular patent claim. That gap between "contributed to the project" and "conceived of the claimed invention" is where most disputes live.

Here's what makes it worse: most sponsored research agreements (SRAs) address ownership without ever clearly addressing inventorship determination. They say something like "jointly made inventions shall be jointly owned" and leave the process of figuring out who actually invented what to be sorted out later. Later, of course, is when the money is real and the relationships are strained.

Joint ownership sounds fair on paper. In practice, it's often a trap. Each co-owner of a U.S. patent can independently license it without the consent of the other owner, and without sharing royalties. That means a company could theoretically license a jointly owned invention to a competitor of the university's spinout, and the university would have limited recourse. Most tech transfer professionals know this. Many industry partners don't, at least not until the moment it matters.

The process problem compounds the legal one. When a potentially patentable invention surfaces, who interviews the inventors? At universities, that's usually the tech transfer office or outside patent counsel. At the company, it's their IP team. These conversations happen separately, with different people asking different questions, using slightly different framings. By the time both sides compare notes, they've sometimes reached different conclusions about who contributed what. Neither side is necessarily lying. Human memory is reconstructive, and conception of an invention is genuinely difficult to pin down after the fact.

Preventable? Yes, mostly. The interventions aren't complicated, but they require doing the work before a dispute exists.

graph TD
    A[SRA Signed] --> B(Joint Research Begins)
    B --> C{Invention Emerges}
    C --> D[Joint Inventorship Review]
    D --> E(Agreed Inventorship Determination)
    E --> F[Patent Filed with Clear Ownership]
    F --> G(License or Spinout Proceeds)
    C --> H[No Process in Place]
    H --> I((Dispute))
    I --> J[Relationship Damage]

First, SRAs should include an explicit inventorship determination process, not just ownership language. Name who conducts interviews, who reviews the analysis, and what happens when the two sides disagree. Mediation provisions cost almost nothing to include at the drafting stage and become invaluable at the dispute stage.

Second, contemporaneous documentation matters enormously. Lab notebooks, dated emails, meeting notes that capture who proposed what: these are what patent counsel actually use to reconstruct conception. Researchers on both sides should understand this. The industry partner's employees should be keeping records that are as rigorous as the academic lab's. Often they're not.

Third, get outside patent counsel involved early, when the invention is still taking shape. Waiting until the invention disclosure is filed means reconstruction is already happening. Bringing counsel in while the research is live allows for real-time documentation and cleaner inventorship analysis.

Fourth, ownership fallback provisions should address joint ownership's specific pathologies. If you're going to allow joint ownership as an outcome, write provisions that restrict unilateral licensing, require revenue sharing, and give each party a right of first negotiation before the other licenses to a third party. Standard joint ownership under 35 U.S.C. § 262 gives you almost none of those protections automatically.

None of this is exotic. Tech transfer professionals who work on federal lab agreements deal with these issues routinely. The gap is that many university-industry SRAs are still using boilerplate that was drafted before co-invention was understood as the liability it actually is.

The relationship damage from a co-invention dispute isn't just about the specific patent in question. Companies that go through a painful dispute with a university tend to route future research to different institutions, or structure future agreements to minimize joint research exposure entirely. That means fewer collaborations, less sponsored funding, and fewer commercialization opportunities down the line. The cost is diffuse and doesn't show up on any spreadsheet, which is exactly why it keeps happening.