By Christopher R. McElwain
The default rule in both copyright and patent law is that the individual who actually creates something (an expressive work in the case of copyright; an invention in the case of patent) owns it. Of course, it’s often the case that the company paying the bills wants to own the IP generated by the folks they’re paying (let’s call them “workers” as shorthand, since, as I’ll discuss below, “employee” is a loaded label). In such a case, it’s common for inventive or creative workers to sign an agreement transferring their rights to their employer. This is called an “assignment.”
Assignments are straightforward when the thing you’re assigning already exists. It works just like a deed or any other document transferring ownership over a piece of property. The contractual pitfalls open up when a worker is asked to hand over the rights to anything and everything they will or might create in the future during their employment. Such prospective assignments are common practice, but they are open to questions about their efficacy, their scope and their timing.
The problem with prospective patent assignments received recent attention due to the Supreme Court’s decision inStanford v. Roche. The Stanford case involved dueling assignment clauses. When Stanford University hired HIV researcher Dr. Mark Holodniy, he signed a contract in which he “agree[d] to assign” his “right, title and interest” in all inventions resulting from his research there. Later, as a condition of gaining access to certain resources at a private medical research company called Cetus, Dr. Holodniy signed another contract agreeing that he “will assign and do[es] hereby assign” his “right, title and interest in each of the ideas, inventions and improvements” made “as a consequence of [his] access.” When Holodniy’s work led to a groundbreaking new HIV blood test, both Stanford and Cetus claimed rights to it.
According to the Federal Circuit court which reviewed the case, Cetus’s “hereby assign” language created a property interest that vested automatically when the blood test was invented, whereas Stanford’s “agree to assign” language did not. Therefore, the later agreement trumped the earlier.
The court’s reasoning followed another Federal Circuit decision from 1991, FilmTec v. Allied-Signal, in which an assignment provision similar to Cetus’s (“. . . agrees to grant and does hereby grant . . .”) was found to transfer legal title “[o]nce the invention is made.” The FilmTec court then contrasted such an assignment with agreements that “merely obligate [a party] to grant future rights.”
Prior to FilmTec, there was a long-held common-law understanding that prospective assignments—whether or not they use the magic word “hereby”—at most give the assignee an equitable interest in the assignor’s hypothetical inventions. To transfer legal title, a written assignment of the particular property was still necessary (a formality that remains in most companies’ IP-best-practices playbooks). The Supreme Court, reviewing the Federal Circuit’s decision in the Stanford case, did not examine the Federal Circuit’s reliance on FilmTec, granting review only on a separate issue, but several of the justices expressed concerns with the reasoning of FilmTec, and Justice Sotomayor wrote a separate concurrence to specifically explicate her view that the majority opinion “permit[s] consideration of these arguments in a future case.” As a result, the future of efficacy of FilmTec-type assignments is unclear.
It’s also unclear whether FilmTec applies to assignments of copyright as well as patents. If an agreement to “hereby assign” an worker’s future inventions becomes effective the moment the invention occurs, why shouldn’t the same be equally true for works that a worker creates subsequent to hereby-assigning away his rights? However, there are important differences between how the Patent Act and the Copyright Act allocate ownership.
Unlike patents, copyrights in works created by true employees are owned by the employer, assignment or none. Works of actual employees created as part of their actual employment are called “Works Made For Hire,” and the employer is considered the author, not the assignee, of such works. Independent contractors, though, are another matter, and the difference between an employee on the one hand and an independent contractor on the other is an involved legal and factual question that has little to do with what the person’s called.
Confusingly, it has become common practice for both employees and independent contractors to sign what’s called a “Work-for-Hire Agreement,” whereby the parties agree to treat the workers’ creations as Works Made for Hire. Since the definition of Works Made for Hire are set by statute, these agreements are unnecessary if the worker is an employee and (in most cases) ineffective if the worker is not. Because most lawyers understand that the statutory definition of a Work Made For Hire can’t be altered by private contract, Work-for-Hire Agreements generally include a fallback assignment clause: “If for some reason the things I create turn out not to be Works Made For Hire, I hereby assign my rights in such works to Company.”
But does this hereby-assignment take effect as soon as the work is fixed (the photograph snapped, the blueprint drafted, the line of code typed) like the FilmTec court said happens with patents? In contrast to patents, the Copyright Act’s Work-for-Hire scheme reduces the burden on employers of obtaining written assignments of specific created works, arguably reducing the need for blanket pre-creation assignments to vest legal title automatically. In addition, copyrights are different from patents in terms of length, scope and subject matter, as well as how the respective rights are obtained, all of which may be reasons to treat them differently when it comes to transfers of not-yet-existent property.
Because of the uncertainties with future assignments, it remains good practice to obtain a second assignment perfecting the transfer once the particular invention has been made or work created. Still, the question of when exactly an assignment occurs will continue to arise. The issue is not a purely academic one. It impacts many common IP scenarios, from assessing damages in cases of infringement to authors exercising their termination rights under the Copyright Act to establishing priority of ownership in cases like the Stanford/Cetus blood-test debacle