The $17 Million NDA That Failed: What IMD v. Cornell Teaches Founders About Patent Protection

Authored by Babak Akhlaghi on June 7, 2026.  A medical device company had an NDA. They had a jury award of more than $17 million. They had a permanent injunction against their competitor.

The Federal Circuit reversed almost all of it.

In April 2026, the court handed down IMD v. Cornell, a case that should make every founder rethink how they protect their technology before sharing it with anyone.

This wasn’t about a defective NDA. The agreement had standard confidentiality language. The problem was upstream: no patentability search, no provisional filing, no execution discipline. The plaintiff walked the defendant right through the NDA’s own escape hatch by asserting information that was already public.

Here’s what happened, why it matters, and what you need to do differently.

What Happened: From Training Session to Federal Circuit

Dr. James Elist founded IMD, a medical device company that developed penile implants. Starting in 2018, his product was the only commercially available instrument for penile enhancement surgery.

In 2019, Dr. Brian Cornell attended a training workshop with Dr. Elist to learn the surgical technique. As part of that training, Dr. Cornell signed an NDA preventing him from disclosing any confidential information he obtained from Dr. Elist.

Later, Dr. Cornell filed his own patents on improvements to the surgical procedure and launched a competing product.

IMD sued for trade secret misappropriation, breach of contract, and patent invalidity (arguing Dr. Elist should have been named as an inventor). A jury awarded more than $17 million in trade secret and exemplary damages, plus $1 million in statutory damages for counterfeiting. The district court entered a permanent injunction.

Dr. Cornell appealed to the Federal Circuit.

The appellate court reversed almost everything. The trade secret verdict collapsed. The inventorship claim failed. The breach of NDA also failed.  Only the trademark counterfeiting award remained standing.

Why the NDA Failed: The “Generally Known” Trapdoor

The NDA wasn’t defective in the abstract. It had a standard provision excluding information that “was or becomes generally available to the public” without Dr. Cornell’s fault.

That’s common language. Most NDAs have it.

The problem was what Dr. Elist chose to assert as confidential and as trade secrets.

Three of the four alleged trade secrets were already disclosed in publicly available patents.

U.S. Patent No. 5,088,477 (issued in 1992 to Subrini) and U.S. Patent No. 4,204,530 (issued in 1980 to Finney) contained the very improvements Dr. Elist claimed as confidential. The Federal Circuit held that “what is disclosed in a patent is ‘generally known to the public’ and cannot be a trade secret.”

Dr. Elist didn’t know about these patents. But the court didn’t care. Ignorance doesn’t matter under the law.

Dr. Elist argued that even if the information appeared in patents, it wasn’t truly “generally known” because those prior patents were never commercialized. The court rejected this completely. Once an idea is disclosed in a published patent application, it becomes generally known, regardless of whether anyone ever builds a product around it.

Publication equals public knowledge. Full stop.

The NDA’s exclusion for “generally available” information became a trapdoor. Dr. Elist walked the defendant right through it.

The Execution Failure: A Wishy-Washy Email

The fourth alleged trade secret was a list of surgical instruments and materials. This one wasn’t in the prior art patents.

It should have been protectable.

But Dr. Elist emailed the list to Dr. Cornell and to a third party (SCA Surgery) without proper confidentiality markings. The email included general confidentiality boilerplate at the bottom, but the language was vague: “may or may not contain confidential information.”

The Federal Circuit found that IMD did not preserve the secrecy of the list’s contents. A wishy-washy email or confidentiality boilerplate that may or may not attach to your attachments does not protect a confidential attachment.

Two fatal flaws:

• The email didn’t specifically attach confidentiality to the list. It hedged on its own protection.
• The list was shared with third parties outside the NDA. If you don’t treat information as confidential, the law won’t either.  And, more importantly, the disclosure of confidential information to third parties who are not obligated to maintain its confidentiality, extinguish the confidentiality and trade secrets protection.

This wasn’t a startup making a rookie mistake. This was an established medical device company with a real product, real IP, and real money on the line. And they still sent that email.

The Inventorship Claim Collapsed Too

Dr. Elist also argued that Dr. Cornell’s patents should be invalid for failing to name him as an inventor. The claim was that the information Dr. Elist provided made him a co-inventor.

But Dr. Elist conceded that his contribution was limited to the trade secret information, and the Federal Circuit had already ruled that information was generally known.

The standard for inventorship requires a significant contribution to the conception of at least one claim. If all your contribution is what’s generally known in the art, that doesn’t amount to significant contribution.

Same facts, same fatal flaw. The trade secret failure killed the inventorship argument too.

File Your Provisional First

Here’s the real lesson: File your provisional application before any disclosure.

Before the investor meeting. Before the partner training. Before the licensing discussion. Before you hand anyone an NDA.

A provisional patent application costs as little as $60 for micro entities (most solo inventors and early-stage startups), $120 for small entities (companies with fewer than 500 employees), and $130 for regular entities. Legal fees for drafting a provisional typically run a few thousand dollars, compared to $10,000 to $20,000 for a non-provisional application.

You get 12 months to test the market, refine the invention, talk to investors, and secure funding before filing a full non-provisional application. You maintain “patent pending” status. You establish a priority date that locks in your place in line.

And here’s the strategic shift: If someone violates the NDA, you’re not chasing a contract claim years later. You’re coming at them with willful patent infringement.

That’s a completely different fight.

Patent infringement is generally a strict liability tort, intent or knowledge is not required to establish liability. By contrast, trade secret misappropriation typically requires proof of improper acquisition, disclosure, or use of information that is not generally known or readily ascertainable.

Patent law focuses on whether an invention is novel and nonobvious over the prior art, while trade secret law focuses on secrecy and commercial confidentiality. As a result, information may be unprotectable as a trade secret because it is generally known, yet still potentially patentable if the specific claimed invention is novel and nonobvious. Conversely, information that is not patentable, such as an incremental improvement or know-how that would be considered obvious, may still qualify for trade secret protection if it provides economic value from remaining secret and is subject to reasonable secrecy measures.

Patent law also permits enhanced damages for willful infringement, while trade secret law may allow recovery based on the defendant’s unjust enrichment or avoided development costs.

The conversation with Dr. Cornell would have been different with a patent in hand. Not a trade secret dispute, a patent infringement claim.

The Prior Art Trap

Dr. Elist may have genuinely believed his improvements were novel and confidential. The record showed the opposite.

How does a founder avoid that trap?

Do a patentability search before you assert anything as a trade secret.

A patentability assessment reviews prior art, publications, and existing patents to determine whether your idea is truly novel or already in the public domain. This should be your first step before deciding whether to pursue patent protection or maintain something as a trade secret.

Not all ideas belong in the patenting bucket. Some fall into trade secrets. Some fall into defensive publications. But you need to know which bucket your idea belongs in before you start sharing it with people.

Talk to an IP attorney. Mention the facts. If patenting is the right approach, file a quick provisional, but a strategic provisional. Make sure it covers the core aspects of your invention, the aspects that competitors care about.

The Real Cost: Market Share

Beyond legal fees, Dr. Elist now has a competitor in his market with a competing product. He has no patent to exclude them.

That’s the real cost. Not the litigation expenses. Not the reversed verdict. At the bottom line, it cost you market share.

I’ve talked to many founders with good ideas validated by the market. A few years down the road, they see competitors entering their space. At that point, they think about patenting. But it’s too late.

As soon as you start selling a product, as soon as you start publicly disclosing the idea to other people, you trigger a time bar against filing your application. Once that time bar expires, the idea is dedicated to the public. You’re left with no choice but to be open to market competition.

The purpose of a patent is to prevent or deter competitors from entering the market. Without one, you’re competing on execution alone.

The Takeaway

The Federal Circuit got this right. For a trade secret claim to survive, the plaintiff has the burden to show that the information was confidential and that the plaintiff derived value from it. If the information was generally known, the claim fails.

The NDA had a built-in escape hatch. Dr. Elist walked the defendant through it by asserting information that was already public.

Here’s what you need to do differently:

• File a provisional patent application before any disclosure. It stops the time bar. It establishes priority. It changes the fight from breach of contract to patent infringement.
• Run a patentability search before asserting trade secrets. Make sure your “secrets” aren’t already in the public domain via patent disclosures.
• Don’t rely on boilerplate confidentiality language. If a document is confidential, label it as such. Make sure your NDA specifically covers all communications and attachments.
• Treat confidential information as confidential. Don’t share it with third parties outside the NDA. Don’t send wishy-washy emails that hedge on their own protection.

You’re not gonna find out about an NDA violation until years down the road. By then, the competitor is in your market, the damage is done, and you’re looking at hundreds of thousands of dollars in litigation costs with no guarantee of recovery.

A provisional patent application costs a few thousand dollars and takes a few weeks to prepare. It’s not a burden. It’s insurance.

File your provisional first. Don’t rely on the NDA alone.