There are alternatives to compulsory invention assignment practice–there’s still a voluntary aspect to nearly all university patent policies, already there, built-in–on the administrative side. There are often two mechanisms in policy by which patent administrators can implement a voluntary program of assignment, even within a policy that makes assignment mandatory:
- decline to assert ownership
- assign back to the inventors if assignment is required by policy.
We have looked at how such a practice can exist within the University of California patent policy. It’s not difficult. It just takes courage to stand up to the folks within the system that think the patent policy is their gift to institutional control, and that somehow that institutional control is the best thing for the diffusion of inventions, for university research, for public benefit, and even for university finances.
Let’s return to the University of Iowa and take a look at how voluntary assignment practice can be implemented there, again without any change in formal patent policy.
University of Iowa
In the case of the University of Iowa, whose patent policy we have used to illustrate various options, the university uses a research foundation, which it controls by placing a majority of university officials on its board. So, in a way, the foundation is also an instrument of the state, but functions as a workaround to state requirements for hiring, contracting, public records, and the like. One might think that such things would be made illegal–that when a university that is an instrument of the state creates “special purpose entities” to move transactions outside state regulatory oversight, then the university should have no involvement in the management of that entity. However, I suppose the operating rule is that university administrators can do any clever thing until the state stops them (the state stopping the state), or gets stopped by the federal government, or by a court–which takes a lawsuit, and money.
Even with a research foundation, the idea of changing practice is straightforward. Even though the patent policy says that the research foundation will “assume” ownership on behalf of the university, there is still a step identified by which inventors must assign ownership to the university:
On behalf of the University, the UIRF shall assume ownership of patents to qualifying inventions. In order for the UIRF to assume ownership, inventors subject to this policy shall assign to the UIRF their entire right in the invention and shall provide reasonable assistance to the UIRF in obtaining patent protection and in licensing the patent rights to others.
The research foundation could simply waive the requirement. There’s a procedure in policy for releasing rights after assignment. But there’s nothing in policy that requires the research foundation to take assignment of every invention. And it makes no administrative sense to take assignment of something that one does not want (for whatever reason, including because the assignment was compulsory without any other reason), simply to release it again immediately. Thus, even at Iowa, with a research foundation, if administrators wanted to test out a voluntary program immediately, all they have to do is to change their practice. Waive the requirement that an inventor assign except where there is a prior agreement with the inventor that he or she has promised to assign (such as for commissioned work, or for work that has been committed under a research agreement or licensing agreement). It is as easy as putting a check box on the disclosure form:
If you have not previously promised to assign this invention to IURF and prefer not to assign the invention to IURF, please check here: [ ]
For disclosures with a checked box, the research foundation would confirm there are no outstanding assignment obligations, and release the inventor from the requirement to assign. The issue is not whether in doing so the research foundation might give up the opportunity to deal in a lucrative invention, but rather whether the research foundation should claim inventions against the will of inventors. That’s a public policy question, not solved by an appeal to institutional hopes for money.
We will get into the value of voluntary assignment. Voluntary assignment restores dual choice invention management, improves relationships between inventors and the invention management agent, allows for a meaningful discussion of strategy and mutual commitments. For now, though, here’s a way to get to voluntary assignment practice by acting within policy rather than changing it. If there is still a matter of equitable interest–perhaps a department has put up significant resources to help the inventor or the inventor has relied her university position to obtain informed consent or to gain access to licensed data sets available only for non-profit work–then there may still be some work to do navigate the situation–but that navigation does not have require assignment of ownership. It could entail providing the university with a financial interest or it could involve agreeing to a covenant that runs with the patent that limits how the patent can be exploited.
We might note that “entire right” is an odd usage in the Iowa patent policy assignment requirement. The broader term would be “entire interest” (see the discussion by Ken Adams). The more specific interest is title to a patentable invention as defined by a patent application that lays out in claims just what has been invented. The main point, however, is that there is an assignment step. The policy here imagines that assignment will always be required and will always take place. But further into the policy, it appears that the university or research foundation (the university in workaround clothing) may waive its right to assignment:
If the UIRF informs in writing the University inventor(s) that it does not wish to file a patent application in any territory based on a disclosure by the inventor(s), the inventor(s) may request from the UIRF an opportunity to take on the prosecution of the patent application. The inventor(s) may request that the UIRF waive its rights to the invention in the territory(ies) in which the UIRF has elected not to file. The UIRF will not unreasonably deny such a request.
The language here is a bit prissy–the foundation has to inform inventors “in writing.” But there is a purpose to it. The effect of this requirement is not to ensure that the foundation does things in a well documented way, but rather just the reverse. If the foundation does not inform the inventors in writing, then they have no right under policy to request a waiver of the foundation’s (university’s) rights. You see how this works? Unless the foundation provides a written notice, it can simply not file patent applications and does not have to respond to any inventor request.
In addition, the inventors–in this clause at least–are constrained if they receive title to the invention. They must file a patent application. While the Iowa patent policy begins with a discourse on the nature of a patent and what inventions qualify for a patent, the policy never gets around to using this information in its discussion of a “qualifying invention.” Nothing in the policy requires an invention to be “patentable,” though it might be inferred that patentable inventions would be the focus. At least the Iowa policy avoids the excesses of definitions of invention that take in non-invention, non-ownable things and treat them as if they were “property” (as, for instance, Wake Forest does with its invention and patent policy).
The upshot of the Iowa approach, then, is that if there’s no patent application filed, then there’s no ownership in the invention to be “assumed” or “waived,” as the case may be. However, given the generality of “entire right” directed to the “invention” and not to a patent application, it is still an open matter whether there is more in play in the “entire right” to an invention than just a right to seek a patent. The looseness is convenient for a university administration doing all the drafting and interpreting of policy. The looseness is less helpful for anyone else trying to make sense of the policy as written.
However, this looseness leads to a misrepresentation in the research foundation’s “model” licensing agreement:
WHEREAS, under the patent policy of The University of Iowa (UI), all inventions and technology arising during the normal course of research and teaching at the UI are assigned and entrusted to the UIRF to obtain patent or other appropriate intellectual property protection and license said technology;
I have highlighted some points. The Iowa patent policy concerns only “inventions” and implies that the inventions of interest are patentable ones. There is nothing about “technology” apart from its involvement in inventions. Further, the patent policy defines qualifying inventions as ones arising (i) in the course of employment; (ii) in a field or discipline “reasonably related” to the field(s) of one’s employment; or (iii) enabled by the “significant use” of University resources.” There is nothing about a “normal course” of anything, nor is the “course of research and teaching” identical to a “course of employment.” One can conduct research in many contexts, only certain of which constitute a “course of employment.” The license template here creates an impression that the university claims ownership of much more than its policy states.
Finally, it boggles the mind to contemplate what other form of “intellectual property protection” besides patent is “obtained” by the activities of an invention administrator obtaining rights under a patent policy. Perhaps what is meant is a plant variety protection certificate or a mask work right. But these are not patentable inventions–they are something else. The implied operation of the patent policy then is this: require assignment of what may be patentable inventions, but if the invention turns out not to be patentable (and therefore not ownable as a matter of patent law), then find some other form of intellectual property to “protect” it and thereby maintain a university ownership claim, even if that ownership was not authorized as claimable by the patent policy and was not assigned by the inventor. Shady? Desperate? Clever?
In general, folks here are imagining a world beyond themselves–what we know (patent) and anything else we can’t bring ourselves to name (other appropriate intellectual property protection). Perhaps they recognize they are clueless and draft about that beyond world just in case it exists.
At any rate, the research foundation won’t “unreasonably” deny a waiver of ownership request–not that addition of the implication that the foundation could come up with reasons not to waive gives any assurance for how the policy plays out in practice. There are other issues with waivers:
However, any waiver of rights will be subject to the interests of any third parties, including, but not limited to, sponsors of the research leading to the invention.
One might wonder what other “third” parties there would be beyond sponsors of research, and why a policy statement is adopting contract language (“third parties,” “including, but not limited to”). What are the parties, here? Where is the contract? If we check the Iowa model research agreement, we see that Iowa grants sponsors a “first option to elect an exclusive license” to “University Intellectual Property developed under this Agreement,” along with a right to sublicense. (Oddly, the model research agreement makes all this dependent on a clause involving joint inventions (8.3) when one would expect the clause that matters concerns the sponsor paying for patent applications on University Intellectual Property (8.4 and 8.5). Perhaps someone will proofread the model agreement before it gets more use).
Here, too, we find inconsistencies with the patent policy. The definition of “University Intellectual Property” in the model research agreement is
“University Intellectual Property” shall mean individually and collectively all inventions, improvements and/or discoveries which are conceived and/or made (i) by one or more employees of University, or (ii) jointly by one or more employees of University and by one or more employees of Sponsor, in performance of the Project.
The university’s patent policy, however, makes claims only with regard to inventions, not “improvements and/or discoveries.” An invention that is only “conceived” is not patentable–it must also be reduced to practice to be “made.” So the the definition of University Intellectual Property here is broader than that of the university’s patent policy.
Furthermore, both the patent policy and research agreement would better be focused in their definitions of invention on “inventions that are or may be patentable.” Otherwise, both the policy and research agreement need to adopt a broader scope of interest, as, say, NASA does with its “Reportable Item” definition in its New Technology FAR Supplement clause (1852.227-70):
“Reportable item” means any invention, discovery, improvement, or innovation of the contractor, whether or not patentable or otherwise protectable under Title 35 of the United States Code, made in the performance of any work under any NASA contract or in the performance of any work that is reimbursable under any clause in any NASA contract providing for reimbursement of costs incurred before the effective date of the contract. Reportable items include, but are not limited to, new processes, machines, manufactures, and compositions of matter, and improvements to, or new applications of, existing processes, machines, manufactures, and compositions of matter. Reportable items also include new computer programs, and improvements to, or new applications of, existing computer programs, whether or not copyrightable or otherwise protectable under Title 17 of the United States Code.
The temptation for university policy drafters, however, is to elevate this broader scoping to the patent policy and make a general claim for all such “new technology” rather than include a statement in the university patent policy that expands the scope only when necessary to meet the requirements of a sponsored research agreement that places claims with such a scope on “new technology.” NASA also uses the “whether or not patentable” construction–but here it for what is reportable, not what NASA claims a right to own. The distinction, however, is lost on most university patent policy drafters, who conflate the scope of what is reportable with what the university claims to own.
Here’s the thing, then, about the university’s reference to “third party” research sponsor rights. If the sponsor declines to pay for patenting, then (apparently, if the cross-references are corrected) the sponsor doesn’t have any rights in University Intellectual Property. If, however, the sponsor does pay for patenting, then there is a patent application on the invention, and the request for waiver provision in the patent policy can’t operate, since the condition for waiver depends on the research foundation deciding that it does not want to a file patent application.
Back to the Iowa patent policy’s waiver procedure:
In addition, any waiver of the institution’s rights in the patent application will expressly allow the University to continue to use the invention for research purposes and will be limited to the scope of the invention as disclosed and as used as a basis for the UIRF’s determination not to file an application in the territory(ies). The UIRF waiver of institutional interest in an invention may result in personal ownership of the invention by University inventor(s) who wish to conduct further research on the invention within the institution. Such inventor(s) should be mindful that use of personally owned patents in an institutional setting may create a conflict of interest requiring disclosure and management under the institution’s policies pertaining to conflict of interest
These are typical conditions that universities place on waived inventions. There is subtlety here, though. The policy is going out of its way to worry that an inventor might not have disclosed all of an invention–and so the waiver is to only what has been disclosed. But the policy does not quite do even this–it uses “scope of the invention as disclosed” not “invention as disclosed.” Thus, the invention is placed in a category of scope without any clarity regarding how such scope is determined. An invention may have possible applications. Must all such applications be disclosed to be waived? An inventor may not have thought of some applications or variations on methods of implementing an invention (a new best mode, say). Are such applications or variations outside the scope of the invention as disclosed and therefore outside the scope of the waiver?
The reservation of rights in waivers of inventions fails to consider shop rights. Universities have a shop right that they enjoy in inventions made with their resources. The demand for rights that Iowa makes, however, is broader than a shop right to the extent that the university claims inventions broader than those made in the course of employment or by employees using university resources.
This isn’t the place to sort out why owning an invention and doing research at the university using it gives rise to a conflict of interest any more than owning a copyright and doing research at the university using it would. But it should be the case that there is no conflict of interest that arises from owning personally an invention or a patent right in an invention. Any conflict of interest that might come about would arise from how the property right is exploited. For instance, if a patent is made available on a FRAND basis to all, then where then is the conflict between public purpose and personal ownership?
What is clear, however, is that university administrators can waive the university’s interest in inventions simply by deciding not to pursue a patent application, provided the inventors agree to pursue a patent application. As with the twisted University of California policy, Iowa makes patenting a condition of release. Once a potential monopoly property right has been recognized, it appears that Iowa insists that the monopoly property right be secured. There is no recognition that inventors might want ownership of an invention to prevent a patent application from being filed (to place the invention in the public domain, or to hold the invention as a trade secret). The complication is, however, that if the inventors add anything to their application that’s not represented in their disclosure to the university, then they are back subject to a potential university claim of ownership in whatever they have added. So to keep things voluntary, anything new in an application gets disclosed and waived in the same manner as the original disclosure of invention.
If the Iowa policy is read to demand assignment, then waiver becomes “re-assignment”–title transfers, and is immediately transferred back to the inventors. Duty done. Policy satisfied. What happens next is based on mutual choice and negotiated agreement. If the university has a financial interest based on its contribution of resources, that’s a matter of equitable sharing of an upside, not ownership of the invention or control of the opportunity. Sharing proceeds, once the agreement is in place, is a much simpler, less expensive proposition than is managing patents and trying to make money from them.