Invention disclosures: what, when and how

Core idea#

An invention disclosure is a confidential document you send to your institution's tech transfer office (TTO) describing something you've invented. It isn't a patent application and it isn't public. It's the formal start of the process by which the institution decides whether to protect the invention, and it often starts legal clocks.

For a researcher who wants to found a company, the disclosure matters more than it looks:

• It records what you invented and when, which protects your priority and inventorship.
• It triggers the institution's obligations (for example, to federal funders in the US) and its decision on whether to file a patent.
• It becomes the basis of the patent application, so a vague disclosure leads to weak claims.
• It's the first step towards a licence to your company (playbooks 2.03, 2.07). A TTO can't license what it doesn't know about.
• Not disclosing can breach your employment terms and cloud the company's IP in due diligence.

The rule of thumb: disclose early, disclose completely, and disclose before any public talk, paper, poster or preprint.

Who it's for: researchers at universities, institutes and government labs whose work might become the basis of a company.

Rules and deadlines#

United States: Bayh-Dole sets the clock for federally funded work#

If the invention was conceived or first reduced to practice with federal funding, the BayhDole Act and its standard patent rights clause (37 CFR 401.14) apply:

1. Two months: the institution must disclose the invention to the funding agency within two months after the inventor discloses it in writing to the institution's patent personnel (37 CFR 401.14).
2. Two years: the institution must elect in writing whether to retain title within two years of disclosing to the agency (37 CFR 401.14). If a publication, sale or public use has already started the one-year US statutory period, that window can shrink to as little as 60 days before the period ends (CASRAI).
3. One year: once title is elected, the institution must file an initial patent application within a year, or earlier if a statutory bar applies (CASRAI).

If the institution misses the disclosure or election deadlines, the agency may take ownership (Fox Rothschild). Reports go through the iEdison system (NIH). A late public disclosure doesn't just risk the patent; it compresses the university's decision time (CASRAI).

Most university IP policies also require disclosure of any invention made with university resources, whatever the funding.

United Kingdom#

There's no Bayh-Dole equivalent, but university employment contracts and IP policies require staff to disclose inventions, and section 39 of the Patents Act usually gives the employer inventions made in the course of normal duties (playbooks 1.15, 2.01). Research council and Innovate UK grants have their own conditions on exploitation and reporting. Students are often treated differently from staff; check your university's policy.

India#

Institute IP policies (IITs, IISc, CSIR labs and others) generally require faculty, staff and students using institute resources to disclose inventions to the IP cell or TTO before any publication. Government-funded projects (for example, DST, DBT or MeitY grants) can carry their own IP and reporting conditions. Under India's absolute novelty standard, disclosing to the TTO first and filing before publishing is essential (playbooks 1.13, 2.04).

Writing a strong disclosure#

Most TTO forms ask for the same things. A strong disclosure is specific, complete and written with the eventual patent and licence in mind.

Inventorship is not authorship#

An inventor is someone who contributed to the conception of at least one claim in the patent application, not everyone who worked on the project or co-authored the paper (CASRAI). Someone who only followed another person's instructions isn't an inventor (University of Toledo). Getting it wrong can invalidate a patent (University of Oklahoma). List everyone who might have contributed to conception, with a note on what they contributed; the patent attorney decides.

For founders, inventorship matters because it often determines ownership: a co-inventor at another institution means that institution may co-own the patent, and you'll need its agreement too (Sterne Kessler).

Worked example (illustrative)#

A postdoc, Li, developed a new membrane coating with her PI and a visiting student from another university. Her first draft disclosure listed all five co-authors of the upcoming paper as inventors and described only the one coating chemistry they'd tested.

After a session with the TTO, she rewrote it:

• Inventors: Li and the visiting student (who proposed the crosslinking step); the PI had funded and supervised but not contributed to conception; two technicians had followed protocols. The patent attorney confirmed this.
• Variations: she added three alternative chemistries and two substrate types she believed would work, with reasoning.
• Applications: water treatment (her target), plus gas separation and battery separators.
• Disclosures: a conference abstract due in seven weeks.
• Third parties: a federal grant and a material transfer agreement for one polymer.

The TTO filed a provisional before the abstract was published, and opened talks with the visiting student's university on a joint IP agreement, early enough to make a licence to Li's company simpler.

Timing and strategy for founders#

When to disclose#

• As soon as you have a definite idea of how it works, not when the prototype is finished. Conception, not reduction to practice, is what matters for inventorship.
• At least four to eight weeks before any public disclosure (talk, poster, paper submission that will be public, preprint, thesis, demo). Patent attorneys need time to draft a provisional.
• When you make a significant improvement. Each improvement may need its own disclosure.

Things founders should do differently#

1. Describe the invention broadly. Include the variations and applications your company might want later, not just what's in the paper.
2. Say you want to found a company. TTOs prioritise inventions with an interested licensee, and some offer founders options or faster terms (2.03).
3. Map third-party rights now. Sponsored research, industry partners, material transfer agreements and co-inventors at other institutions all affect what can be licensed.
4. Keep a dated record of your work (lab notebook, version control). It supports both the disclosure and later disputes.
5. Ask what happens next, and when: the TTO's assessment, a decision to file, the provisional filing date, and when you can talk publicly.

If the TTO decides not to file#

Some universities will return or release rights to inventors if they don't pursue a patent; Bayh-Dole also allows agencies to let inventors retain title in some cases. Ask about the policy. A released invention may still be worth building a company on, especially if it can be protected as a trade secret (2.18) or through later patents.

After disclosure#

• Don't present the invention publicly until the TTO confirms a filing.
• Use NDAs for conversations with companies and investors before filing (and remember many investors won't sign them, so share outcomes, not methods).
• Update the disclosure if new data or variations emerge before the full application is drafted.

Scripts#

Sending the disclosure#

"Please find attached an invention disclosure for [title]. We have a conference abstract due on [date], so we'd be grateful for a view on filing before then. I'm also interested in forming a company to commercialise this and would welcome a conversation about an option or licence."

Asking the TTO about next steps#

"Could you tell me the expected timeline: when you'll decide whether to file, when a provisional could be in place, and when I'll be free to present? And are there any third-party rights you've spotted that we should resolve early?"

Raising inventorship with a colleague#

"The TTO asked us to list everyone who contributed to the conception of the invention, which is narrower than authorship. I've listed you for [contribution]. The patent attorney makes the final call. Does that match your understanding?"

Common mistakes#

• Disclosing after the talk or paper. It can destroy patent rights outside the US and compresses deadlines inside it.
• Copying the paper into the form. Papers are written to show one result; disclosures need variations and applications.
• Treating the author list as the inventor list.
• Forgetting funders, sponsors and material transfer agreements.
• Not mentioning your interest in founding.
• One disclosure for a stream of improvements. Disclose each significant one.
• Talking to investors in detail before filing.

Related playbooks#

• 1.13 Building a public profile (disclosure risks)
• 2.01 Who owns your invention?
• 2.03 Working with the tech transfer office
• 2.04 Patent or publish first?
• 2.05 Provisional filings and the PCT timeline
• 2.13 Students and postdocs as founders

Sources#

• 37 CFR 401.14, standard patent rights clause; CASRAI, 37 CFR 401 explained and invention disclosure basics; NIH, inventions and patents; Fox Rothschild, disclosure deadlines.
• CASRAI, authorship vs inventorship; University of Toledo, determining inventorship; University of Oklahoma, inventors vs non-inventors; Sterne Kessler, understanding inventorship.

Worked example is illustrative. Not legal advice. Last checked November 2026.