Fail-Safe v. A.O. Smith: Trade Secrets

Q: Does sharing information without an NDA automatically destroy a trade secret?

Not by itself, but it is powerful evidence of failure. Fail-Safe holds that a total absence of any confidentiality measures during business discussions can defeat the reasonable-measures element as a matter of law. An NDA is the most common precaution, though not the only one.

Seventh Circuit affirmed summary judgment against a plaintiff that shared technology with no NDA: zero precautions fails the reasonable-measures test.

Trade-secret law asks little of a plaintiff to clear the secrecy threshold, but it does ask for something. Fail-Safe, LLC v. A.O. Smith Corp., 674 F.3d 889 (7th Cir. 2012), decided March 29, 2012, is the decision that shows what happens when a company offers nothing at all. A panel of the United States Court of Appeals for the Seventh Circuit (Chief Judge Easterbrook, Circuit Judge Cudahy, and District Judge Pratt sitting by designation) affirmed summary judgment against a technology developer that had shared sensitive engineering information over four years of joint-venture talks without ever asking for a confidentiality agreement, marking the material as confidential, or so much as raising the subject of secrecy.

The case matters because it treats the “reasonable efforts” element of a trade secret not as a jury-only fact question but as something a court can dispose of on summary judgment when the record shows a plaintiff took no precautions whatsoever. For any business weighing whether to guard its disclosures during a courtship with a larger partner, Fail-Safe is the cautionary tale that a promising invention plus zero protective steps equals no protectable secret.

At a glance

Case: Fail-Safe, LLC v. A.O. Smith Corp., 674 F.3d 889, No. 11-1354 (7th Cir. 2012)
Decided: March 29, 2012; panel of Chief Judge Easterbrook, Circuit Judge Cudahy, and District Judge Pratt (by designation); affirming summary judgment for the defendant
Holding: Under Wisconsin’s Uniform Trade Secrets Act, a plaintiff that takes no measures at all to maintain secrecy fails the reasonable-efforts element as a matter of law, so its information is not a protectable trade secret.
Status: Final. The Seventh Circuit affirmed the district court’s grant of summary judgment.

The reasonable-measures element and why it is rarely dispositive

Wisconsin has adopted the Uniform Trade Secrets Act at Wis. Stat. § 134.90, which mirrors the model act followed by the great majority of states. To qualify as a trade secret, information must satisfy two requirements: it must derive independent economic value from not being generally known, and it must be “the subject of efforts to maintain its secrecy that are reasonable under the circumstances.” The first prong looks at value; the second looks at conduct.

Most reasonable-measures disputes never resolve cleanly. Reasonableness is a classic fact-intensive question, and courts frequently send it to a jury because the sufficiency of a company’s precautions (password rules, restricted access, marked documents, exit interviews, and the like) depends on context. A modest set of safeguards may be reasonable for a small firm and inadequate for a large one. That contextual quality is precisely why defendants rarely win the element outright on summary judgment.

Fail-Safe is the outlier that proves the rule. The Seventh Circuit did not hold that Fail-Safe’s precautions were thin or imperfect. It held that there were none. When the record contains no protective measures at all, the reasonableness inquiry collapses: there is nothing for a jury to weigh, because the plaintiff has offered no conduct to be weighed as reasonable or unreasonable.

The facts: four years of disclosure, no confidentiality

Fail-Safe developed an anti-entrapment pump system designed to prevent swimming-pool drains from trapping and injuring swimmers, a genuine safety problem that drove regulatory attention to pool-drain design. The contact began in November 2000, when A.O. Smith representatives saw Fail-Safe demonstrate its technology at an industry trade show in Orlando, and grew into extended discussions with A.O. Smith Corporation, a large manufacturer of electric motors, about jointly developing and marketing the technology. Over roughly four years of meetings, calls, and correspondence, running until the exchanges stopped in October 2004, Fail-Safe shared substantial technical information about its approach, including letters describing specific features and test results.

What Fail-Safe never did was protect that information. It did not ask A.O. Smith to sign a confidentiality or nondisclosure agreement covering Fail-Safe’s disclosures. It did not stamp its materials confidential. It did not tell A.O. Smith that the information was secret or that it expected the information to be kept in confidence. The subject of confidentiality, by the court’s account, was never raised during the parties’ interactions.

The omission was striking because Fail-Safe knew how to protect itself. The record showed that Fail-Safe had used confidentiality agreements with other potential partners in the past, so this was not a company ignorant of the practice. More pointedly, the relationship did feature a nondisclosure agreement, but it ran the wrong way: Fail-Safe signed A.O. Smith’s one-way NDA protecting A.O. Smith’s information, without ever securing reciprocal protection for its own disclosures. Fail-Safe thus bound itself to keep the larger company’s secrets while leaving its own entirely exposed.

The joint project ultimately fell apart. Fail-Safe later alleged that A.O. Smith incorporated its ideas into A.O. Smith’s own pump-motor products and sued for trade-secret misappropriation. The district court granted summary judgment for A.O. Smith, and Fail-Safe appealed.

The court’s reasoning: no precautions, no secret, as a matter of law

The Seventh Circuit affirmed, and it did so squarely on the reasonable-measures element rather than on the harder questions of value or actual copying. The panel emphasized that Fail-Safe “failed to take any precautionary measures to protect its claimed trade secrets.” That total absence, not merely a shortfall, was what allowed the court to resolve the issue without a trial.

The opinion framed the disclosures as freely given business advice rather than guarded secrets. Turning down Fail-Safe’s fallback unjust-enrichment theory, the court delivered the line practitioners have quoted since: “Simply put, one cannot steal free advice.” The intuition is that a company which hands over information with no strings attached cannot later reclassify that information as a stolen secret. Agreeing with the district court, and borrowing the phrase from the Wisconsin Supreme Court’s decision in RTE Corp. v. Coatings, Inc., 267 N.W.2d 226 (Wis. 1978), the panel said Fail-Safe “courted [its] own disaster” by failing to take any protective measures.

The court treated the surrounding circumstances as aggravating rather than excusing. Fail-Safe was a sophisticated party that had used confidentiality agreements elsewhere, so it could not claim inexperience. And it had actually executed A.O. Smith’s NDA, demonstrating both awareness of the tool and a decision not to deploy it in its own favor. On that record, the panel concluded, no reasonable jury could find that Fail-Safe made reasonable efforts to maintain secrecy, so the information was not a trade secret and the misappropriation claim failed as a matter of law.

Because the reasonable-measures ground was dispositive, the Seventh Circuit did not need to rest its decision on the timeliness questions that had also surfaced below, where the age of the claim raised limitations and laches concerns. The merits ruling on secrecy was enough to end the case.

What Fail-Safe changed, and what it confirmed

Fail-Safe did not announce a new legal standard. The reasonable-efforts requirement long predates it, tracing through the UTSA and older common-law formulations. What the decision contributed was a clean appellate holding that the element can be decided as a matter of law at the zero-precaution extreme, and a vivid factual template of how a company forfeits protection during business negotiations.

The decision sits alongside a line of cases stressing that secrecy is behavioral. Courts have repeatedly held that information does not become a trade secret merely because it is valuable or non-obvious; the owner must act like it is a secret. Fail-Safe pushes that principle to its logical endpoint: when the owner acts in every respect as though the information is not a secret, a court need not defer to a jury to say so.

For litigators, the case is a summary-judgment tool. A defendant facing a misappropriation claim will comb the record for the plaintiff’s precautions and, if it finds none tied to the specific disclosures at issue, will argue Fail-Safe to take the secrecy element away from the jury. For transactional lawyers, the case is a checklist item: the reciprocal NDA that Fail-Safe never obtained is the single step that most likely would have saved the claim.

Open questions

Fail-Safe draws a bright line at the zero-precaution extreme but leaves the middle ground unresolved. How few measures are too few before a court can rule as a matter of law rather than sending the question to a jury? The opinion answers the easy case (nothing at all) without mapping the gradient between “nothing” and “clearly reasonable.”

The decision also does not settle how courts should treat one-way NDAs in mixed relationships. Fail-Safe’s signature on A.O. Smith’s agreement counted against it, but the opinion does not fully explore whether executing a partner’s NDA can ever supply evidence of a shared confidentiality understanding covering one’s own disclosures. Later courts continue to weigh course-of-dealing and implied-confidentiality arguments case by case, and Fail-Safe does not foreclose them where at least some protective conduct exists.

Implications for inventors and businesses

Get the NDA before you talk. The lesson of Fail-Safe is not subtle. Execute a written confidentiality agreement covering your own disclosures before substantive technical discussions begin, not after a relationship has already produced four years of unguarded exchanges.
Make it reciprocal. Signing a larger partner’s one-way NDA protects them, not you. If you will be disclosing your own technology, insist that the agreement runs both directions so that the information you hand over is contractually restricted.
Mark and manage what you share. Beyond the contract, label sensitive materials as confidential, limit who receives them, and keep a record of what was disclosed and under what terms. These steps build the reasonableness case that Fail-Safe could not make.
Assume the secrecy element will be tested first. Defendants increasingly move for summary judgment on reasonable measures rather than fighting over value or copying. Document your precautions contemporaneously so that, unlike Fail-Safe, you have conduct to point to.

Frequently asked questions

Does sharing information without an NDA automatically destroy a trade secret? Not by itself, but it is powerful evidence of failure. Fail-Safe holds that a total absence of any confidentiality measures during business discussions can defeat the reasonable-measures element as a matter of law. An NDA is the most common precaution, though not the only one.

Can a court decide the reasonable-measures question on summary judgment? Yes. While reasonableness is often a jury question, the Seventh Circuit held that where a plaintiff took no precautions at all, a court can resolve the element as a matter of law without sending it to trial.

Why did signing A.O. Smith’s one-way NDA not help Fail-Safe? That agreement protected A.O. Smith’s information, not Fail-Safe’s. Fail-Safe never asked A.O. Smith to keep Fail-Safe’s disclosures confidential, so the information Fail-Safe handed over went out without any contractual restriction.