Last week the United States Supreme Court clarified the respective invention ownership rights between federal contractors and their employee-inventors under the Bayh-Dole Act (35 U.S.C. §§201(e), (c), 202(a)) (the “Act”) in Board of Trustees of the Leland Stanford Junior University v. Roche Molecular Systems, Inc. (Sup. Ct. June 6, 2011). While blandly holding that the Bayh-Dole Act of 1980 does not automatically vest title to federally-funded inventions in federal contractors, the Court left intact a Federal Circuit rule that Justice Breyer in dissent characterized as “a technical drafting trap for the unwary.” The trap is sprung by the slight linguistic differences between the phrases “agree to assign” and “do hereby assign.” This case has implications on compliance with federal contracts and the practice of technology transfer, and should increase the scrutiny given invention assignment agreements between acquisition targets and their employee inventors.

Case Holding

In Stanford v. Roche, the Court analyzed the allocation of rights to “subject inventions,” which are defined under the Act as an “invention of the contractor conceived or first actually reduced to practice in the performance of work under a funding agreement.” Stanford argued the Act divests the contractor’s employees’ rights to their inventions made in the performance of work under a federal funding agreement, and assigns them to the federal contractor-employer, unless they have expressly elected otherwise. The Court disagreed, holding the Act does not contradict legal precedent that allows employee inventors to retain title to their subject inventions absent an express assignment otherwise, and thus concluded that the invention in question was not “an invention of the contractor” and therefore not a “subject invention” to which Stanford could retain title.

Case Analysis

The phrase “agree to assign” has been interpreted by the Federal Circuit as a mere promise to assign rights in the future which requires a subsequent assignment. Board of Trustees of the Leland Stanford Junior University v. Roche Molecular Systems, Inc., 583 F.3d 832, 842 (Fed. Cir. 2009) (citing IpVenture, Inc. v. Prostar Computer, Inc., 503 F.3d 1324, 1327 (Fed. Cir. 2007) as ‘interpreting ‘agree to assign’ as ‘an agreement to assign’ requiring a subsequent written instrument.’). In contrast, the words “do hereby assign” have been construed as a present assignment of future rights. FilmTec Corp. v. Allied-Signal, Inc., 939 F.2d 1568, 1572-73 (Fed. Cir. 1991) (emphasis added). This language transfers title by operation of law, without the need for any subsequent assignment, at such time as the invention is made and the patent application is filed. Under the FilmTec rule, a subsequent assignment of the invention, as ordinarily occurs during patent prosecution, is invalid where a present assignment of future rights exists.

Stanford v. Roche arose when Stanford sued Roche for infringement of three of its patents related to methods for detecting HIV using the polymerase chain reaction (PCR). All three of the patents descended from a common parent application filed in 1992. All of the named inventors on that application are employees of Stanford. In 1988, Mark Holodniy, one of the inventors on the 1992 application, joined Stanford as a research fellow. He signed a “Copyright and Patent Agreement” that obligated him to assign his future inventions to Stanford using the words “I agree to assign.” In 1989, as part of his research to develop a PCR-based method for detecting HIV, Holodniy began a series of visits to Cetus Corp. Holodniy was asked by Cetus to sign a document disarmingly entitled, “Visitors Confidentiality Agreement.” But that agreement contained more than confidentiality provisions. It also said that visitors such as Holodniy “do hereby assign” to Cetus inventions made “as a consequence of” work at Cetus. The result under FilmTec was that Cetus immediately gained equitable title to Holodniy’s later PCR-related inventions. So when Stanford filed the parent patent application in 1992, Cetus’s equitable title converted to legal title in Cetus by operation of law.

The Supreme Court explained that an inventor retains rights in her or his invention unless and until she or he assigns it to another, including inventions that are made in the performance of work under a federal funding agreement and therefore presumed to be “subject inventions” under the Act. The Court agreed with the Federal Circuit’s reasoning that the future tense language in the Stanford agreement (“agree[d] to assign”) was trumped by the present-tense language in the Cetus agreement (“do[es] hereby assign”). Therefore, because Holodniy did not assign his inventorship rights to Stanford, he retained title to those rights (which he then assigned to Cetus), and the Act did not demand otherwise. The Act did not alter this allocation of rights because, the Court held, they did not become “subject inventions” unless and until they first became inventions of Stanford.

Stanford also attempted to overcome its defective chain of title by claiming it was a “bona fide purchaser” of Holodniy’s interest in good faith and without notice of the prior assignment to Cetus. The Federal Circuit charged Stanford with at least constructive or inquiry notice of the Cetus agreement, noting that an organization can be charged with notice of its employees’ assignments. And the fact that Holodniy’s Cetus agreement violated the terms of his earlier agreement with Stanford, which provided that he would not enter into any agreement creating conflicting patent obligations, did not prevent the Federal Circuit from imputing notice of the agreement to Stanford.

Concerns for Government Contractors and Grant Recipients

A primary purpose of the Act is to encourage institutions that receive federal research funding to commercialize their subject inventions for the public benefit. The institution may choose to assign title to subject inventions to the government. If the institution elects to retain title, however, then it must grant the government a nonexclusive license to practice the invention. If the institution retains title but does not patent and commercialize the subject invention, then the government may “march in” and license the subject invention to others. Terms of federal contracts and grants arising from other statutes also may require the contractor or grant recipient to provide certain rights to the government or to third parties. Contractors and grant recipients have frequently assumed, as did Stanford, that employee inventions conceived or reduced to practice on government funded work provided the basis for complying with these requirements.

This decision illuminates the gap between the law governing ownership of inventions and the duty of the government contractor with respect to inventions that are “subject inventions” under the Act. The Act does not impose upon the contractor a duty to obtain ownership of its employees’ inventions. But if a contractor fails to take assignment of its employees’ inventions (as the Court held Stanford failed to do), then the employee is free to retain those rights, and may even license or assign those rights to others (as Holodniy assigned his rights to Cetus). This may lead to the contractor breaching its duties to the Government under the contract or grant.

Employment Agreements and Due Diligence

This trap is not limited to federal contractors under Bayh-Dole. As this case illustrates, an employer’s lack of actual knowledge of its employee’s conflicting agreement, even when that agreement amounted to a breach of duty to the employer, is insufficient to avoid the trap. And, even where, as here, a third party may be blocked from asserting ownership due to the running of the statute of limitations, it will not prevent a defense based on lack of standing. This can be fatal to an infringement action, as it was to Stanford here, because all co-owners must join as plaintiffs in such an action.

Stanford v. Roche also illustrates the threat to a contractor’s ownership of, and ability to commercialize the results of, research performed using federal funds that are not subject to correct and current assignment agreements from the inventing employee. Anyone acquiring a company whose agreements with its inventor-employees have not been correctly drafted may find itself without the title to patents that it thought it was acquiring. Due diligence of acquisition targets needs to uncover this set of risks.

Justice Breyer’s dissent criticizes the court’s reliance on the distinction between the language of the Cetus and Stanford agreements as a drafting trap. Nevertheless, the contractor can no longer assume that ownership of its employees’ inventions automatically vests with it by operation of the Act, and must be very careful in drafting their employment agreements. Contractors should also consider evaluating their existing agreements to confirm ownership and freedom to commercialize research conducted under the Act.