The IP Assignment Agreement You Didn't Read Will End Your Startup

Most technical founders think the hardest IP problem they'll face is writing strong patent claims. It isn't. The hardest problem is figuring out whether you actually own what you built — before a Series A lawyer does it for you, in a due diligence memo, three weeks before a term sheet expires.

Photo by RDNE Stock project on Pexels.

Here's the scenario: You spent four years developing a novel biosensor in a university lab. You filed a provisional with the tech transfer office. You licensed it back for a modest royalty. You raised a seed round. Everything feels legitimate.

Then the acquirer's counsel asks a simple question: "Did any of the inventive work happen on university equipment, under a federally funded grant, or during hours when you were employed as a researcher?"

Of course it did. All of it.

What follows is rarely a clean answer.

The Document You Signed on Day One

Every university employee and most graduate students sign an IP assignment agreement before they ever set foot in a lab. These agreements vary — some cover only work done with institutional resources, others claim anything you invent that's "related to" your research area, full stop. Read yours. Right now, if you haven't.

If you can't find it, file a public records request or ask HR. Do not assume you know what it says based on what a colleague told you over coffee.

The dangerous assumption is that spinning out through the tech transfer office resolves all of this. It doesn't. A license gives you the right to use the IP; it doesn't mean the ownership question is clean. Investors want to see clear title, not a royalty agreement that a university can modify or terminate if you miss a milestone.

Federal Funding Makes Everything More Complicated

If your research was funded by NSF, NIH, DOE, or any other federal agency, the Bayh-Dole Act governs what happens to inventions that came out of it. Broadly, the university can elect to retain title — and usually does — but the government retains a license and certain "march-in" rights.

March-in rights sound theoretical until they aren't. The government can, in principle, require the patent holder to license the technology to third parties if they determine it's not being adequately commercialized. This has rarely been exercised, but investors know it exists, and some will ask about it.

More practically: if your startup's core asset is a patent that has a federal funding disclosure obligation and that disclosure wasn't made, the patent can be invalidated. This has happened. It will happen again.

graph TD
    A[Invention Created] --> B{University Resources or Federal Funding?}
    B -- Yes --> C[University Claims Ownership]
    B -- No --> D[Founder May Own Directly]
    C --> E{Tech Transfer Election}
    E -- Retained --> F[License Back to Startup]
    E -- Abandoned --> G[Founder Can Potentially Acquire]
    F --> H[Investor Due Diligence]
    D --> H

What "Scrubbing" the IP Actually Means

Some founders try to recreate their core technology outside the university after leaving, using no institutional equipment and no federal dollars. This is called scrubbing, and it can work — but only if you do it correctly and document everything.

Correctly means: new lab notebooks, new dates, new materials purchased with personal funds, and ideally a gap in time between your university affiliation ending and your independent development beginning. A weekend doesn't cut it.

Documentation means: timestamped files, purchase receipts, witness signatures on notebook pages. Courts have examined this evidence. So have patent examiners.

If you can't scrub cleanly — because the invention is too mature, too dependent on earlier work, or because the university has already filed — stop trying to obscure the chain of title. Work with a real IP attorney to either negotiate a clean license or acquire the patent outright.

The Conversation You're Avoiding

Founders delay talking to IP counsel for the same reason they delay talking to a doctor about a weird mole: they'd rather not know. But the cost of resolving a messy ownership question in year one is a fraction of what it costs during a financing round or acquisition.

Get an IP freedom-to-operate opinion. Have someone map every claim in your core patents to the work you did, where you did it, and who funded it. If there are gaps or risks, document them and disclose them proactively.

Investors have seen uglier things. What they can't forgive is a founder who knew about a problem and buried it.

Ownership isn't just a legal formality — it's the foundation everything else is built on. If that foundation has cracks, find them yourself before someone else does.