When the Invention Can't Work

Q: Can you patent a cold-fusion device?

Only if you can prove it actually works. In re Swartz holds that because published references show cold-fusion results are irreproducible, the Patent Office may reasonably doubt the operability of such an invention, shifting the burden to the applicant to provide evidence sufficient to convince a skilled person that the claimed process is operable.

Q: How does utility relate to enablement under Section 112?

They are linked. Swartz holds that if claims fail the utility requirement because the invention is inoperative, they also fail the enablement requirement of 35 U.S.C. § 112, because a person skilled in the art cannot practice — cannot make and use — an invention that does not work.

The Federal Circuit affirmed the rejection of a cold-fusion patent application, holding that once the Patent Office shows skilled artisans would reasonably doubt an invention's utility, the burden shifts to the applicant to prove it works.

Patent law presumes that inventors tell the truth about what their inventions do, but that presumption has limits — and few inventions test those limits like cold fusion. In In re Swartz, 232 F.3d 862 (Fed. Cir. 2000), the Federal Circuit affirmed the rejection of Mitchell Swartz’s application for a cold-fusion process, holding that the Patent Office had properly doubted its operability and that Swartz had failed to prove the invention actually worked. Decided per curiam by a panel of Judges Plager, Archer (Senior Circuit Judge), and Dyk, the short opinion has become a workhorse citation for the proposition that an invention which cannot operate as claimed is unpatentable — and that “incredible utility” is a legitimate ground of rejection.

At a glance

Case: In re Swartz, 232 F.3d 862 (Fed. Cir. 2000) (per curiam).
Court: U.S. Court of Appeals for the Federal Circuit; panel of Circuit Judge Plager, Senior Circuit Judge Archer, and Circuit Judge Dyk.
Posture: Appeal from the Board of Patent Appeals and Interferences, which had affirmed the examiner’s rejection of the claims in Application Serial No. 07/760,970 directed to a cold-fusion process.
Holding: Where the Patent Office shows that one of ordinary skill in the art would reasonably doubt an invention’s asserted utility, the burden shifts to the applicant to prove operability; cold-fusion claims fail for lack of utility under 35 U.S.C. § 101 and, derivatively, enablement under § 112.
Significance: A leading modern authority on operability, “incredible utility,” and the tie between the utility and enablement requirements.

The application and the rejection

Mitchell Swartz sought a patent on a process said to produce energy through cold fusion — nuclear fusion at or near room temperature, the phenomenon famously announced and then widely doubted after the 1989 Pons-Fleischmann claims. The examiner rejected the claims under 35 U.S.C. § 101 for lack of utility and operability and under § 112, paragraph 1, for lack of enablement, reasoning that the asserted invention did not work as claimed. The Board affirmed, and Swartz appealed to the Federal Circuit.

The factual record drove the result. The examiner relied on several references documenting that experimental results in the cold-fusion field were irreproducible — that researchers could not reliably repeat the claimed energy-producing effects. The agency treated that body of literature as evidence that a skilled artisan would reasonably doubt whether Swartz’s process could operate at all. Swartz, for his part, had not supplied the kind of experimental data that would persuade such a person otherwise.

The burden-shifting framework

The opinion restates the governing burden-allocation rule with unusual clarity. An applicant’s assertion of utility is “presumptively correct,” and the Patent Office bears the initial burden of challenging it. But that burden is not heavy where the science is suspect. “If the PTO provides evidence showing that one of ordinary skill in the art would reasonably doubt the asserted utility,” the court held, “the burden shifts to the applicant to submit evidence sufficient to convince such a person of the invention’s asserted utility.”

Applying that framework, the court found that the references on irreproducibility supplied “substantial evidence” that skilled artisans would “reasonably doubt the asserted utility and operability” of cold fusion. Because the agency carried its initial burden, the obligation moved to Swartz to come forward with persuasive proof of operability — and he had not. The court therefore affirmed the § 101 rejection. The framework is even-handed in principle: an ordinary, plausible invention enjoys the presumption of utility and the agency rarely meets its initial burden. It is only the extraordinary claim, contradicted by the scientific record, that triggers the demand for affirmative proof.

Utility, operability, and enablement together

Swartz is cited as often for what it says about enablement as for utility. The court held that the failure of utility carried the enablement rejection with it: “If the claims in an application fail to meet the utility requirement because the invention is inoperative, they also fail to meet the enablement requirement” of § 112, because “a person skilled in the art cannot practice the invention.” The logic is straightforward — one cannot teach the public how to make and use something that does not work in the first place.

That linkage gives examiners a doctrinal one-two punch against inventions that defy established science, from cold fusion to perpetual-motion machines. The decision sits alongside cases like Newman v. Quigg (perpetual motion) and is reflected in the Patent Office’s own utility guidance in the Manual of Patent Examining Procedure. Swartz himself continued to pursue cold-fusion patents for years, and the Federal Circuit again affirmed rejections in later proceedings, underscoring how durable the operability bar has proved. The case stands for a sober principle: enthusiasm and theory are not enough; an applicant claiming a result the scientific community deems unattainable must show, with evidence, that the invention actually operates.

Open questions

How much agency evidence triggers the shift? Swartz requires the PTO to show that a skilled person would “reasonably doubt” utility, but the threshold quantum of evidence for that showing is fact-specific.
Where is the line between unproven and inoperative? Emerging science is often unproven without being impossible; Swartz does not draw a bright line between the two, leaving room for argument as fields mature.
What proof rebuts the doubt? The opinion demands evidence “sufficient to convince” a skilled artisan but does not specify what experimental showing would suffice for an extraordinary claim.

Implications

“Incredible utility” is a real rejection. Inventions that contradict accepted science can be refused for lack of operability without a working-model requirement for ordinary cases.
The burden can shift to the applicant. Once the agency shows reasonable doubt, the inventor must affirmatively prove the invention works — usually with data, not argument.
Utility and enablement rise and fall together. An inoperative invention fails both § 101 and § 112, so a single operability problem can sink an application on two grounds.
Document operability for frontier technologies. Applicants in cutting-edge or contested fields should build a reproducible evidentiary record before filing.
Examiners may invoke the scientific consensus. Published literature showing irreproducibility can meet the agency’s initial burden, making peer-reviewed skepticism a litigation risk at the patent office.

Frequently asked questions

Can you patent a cold-fusion device? Only if you can prove it actually works. In re Swartz holds that because published references show cold-fusion results are irreproducible, the Patent Office may reasonably doubt the operability of such an invention, shifting the burden to the applicant to provide evidence sufficient to convince a skilled person that the claimed process is operable.

What is the utility or operability requirement under Section 101? 35 U.S.C. § 101 requires an invention to be useful. An inventor’s assertion of utility is presumed correct, but if the Patent Office produces evidence that one of ordinary skill in the art would reasonably doubt the asserted utility — for example, because the invention appears inoperative — the applicant must come forward with proof that it works.

How does utility relate to enablement under Section 112? They are linked. Swartz holds that if claims fail the utility requirement because the invention is inoperative, they also fail the enablement requirement of 35 U.S.C. § 112, because a person skilled in the art cannot practice — cannot make and use — an invention that does not work.