This turned out to be a longish essay for a blog environment. It’s not for everyone. It puts together arguments against the idea that present assignments somehow address the Stanford v Roche situation, or situations like it, or are otherwise a really good thing for university IP practice. I explain once again the history of university approaches to inventions, and show how a new policy architecture is being built, badly, on the remains of an older, modestly productive one. I point out the problems in the logic of present assignments and show that there are better alternatives. A key point the essay develops is that present assignments of undocumented assets are a nightmare for any licensing practice in an open research environment that uses any kind of exclusive license or a first right or exclusive right to negotiate.
Over the past few weeks, I have been working on present assignment issues. I have picked apart the present assignments at University of California and University of Washington, as well as the policy-based attempts at the University of Colorado and the University of Texas. The attorneys publishing legal commentary appear uniformly to be on the side of helping universities implement present assignments. The argument being made is that doing so will address the Stanford v Roche situation, improve “title certainty,” and will be generally a good thing for universities to do. I don’t agree. It’s not a superficial not agreeing, but rather one that comes from twenty years of practice, time running technology transfer programs at major universities, a lot of consulting with folks at universities foreign and domestic, and a lot of time spent reading policies, agreements, law cases, and annual reports wherever I have been able to find them.
I won’t claim that practice is always right, or that what I’ve experienced and understand makes it so, but there is something to be said for practice–for being on the field and doing it–that armchair quarterbacks don’t have, and for much of the law profession, when it comes to technology transfer, they only watch the game when the refs have called a penalty. Then they perk up, make massive assumptions about the role of the penalty dispute in the course of the game, and offer authoritative insights into whether the call was wrongly made and what reasoning about the call teaches us about how to get away with things in the future. While all this can be a benefit to practice, it can also obscure the purpose of the game and the ways in which it is played, and can be played.
When the game is innovation sparked by research expertise and discoveries, then more is on the line than a football contest, however. The push for present assignments is driven by a set of misunderstandings about Stanford v Roche, a set of worries about proper administration of patents, and an apparent belief that compulsory university ownership of faculty and student intellectual property is desirable, even necessary. This approach appears to be ignorant of how Bayh-Dole operates, its context for universities as part of Circular A-110 [now 2 CFR 200] grant requirements, and what the purpose of all the requirements are anyway: to situate universities as stewards of research enterprise, give investigators and inventors a degree of choices in the agents that they work with to develop inventions, and motivate everyone involved to get research findings out into practical use.
For all this institutional ownership and money-making from licensing are a side-show, not the primary goal. Bayh-Dole was not the government’s gift to universities as a way to make extra money for research by extorting industry and their own faculty. Yet this is how it is turning out. The amounts for research are woefully small compared to the funds available from sponsors and donors, the amount of ill-will is substantial, even when the tech transfer folks are heroic in their efforts to overcome institutional inertias and reticences and politics, and the effect on personal initiative and individual engagement with the broader community is substantial, but is not going to be documented by surveys or reports of missed opportunities because people were no longer looking for them.
I know, how can one make any arguments without statistics? I agree with Taleb on this one: some things are done by thinking about them, not playing with “metrics.” That is, metrics capture some of what is observed, but learning to observe is not a function of metrics–and may be made more difficult by their emphasis. Remember, we are talking here about discovery and innovation, not running a department store. We are looking for black swans, to create them, to be part of them, not trying to plan them out of existence. Our business is weak rhetorics, emerging signals, rare events, bewilderment. What we do is more likely to give angst than feck to the status quo, but it also opens up the status quo for new opportunities, new leadership, new ways of making and remaking community. Jobs, health, security–and imagination and entertainment–you know, the idealistic stuff once one gets over the idea that it’s all really about institutions getting as much government money as they can while spinning a plausible success story to the public.
So let’s look at what’s wrong with present assignments as they are being implemented by universities. I’m arguing that the legal advisors have got it wrong, that they don’t appear to know university technology transfer policy and practice, and their advice is damaging to universities, to research and scholarship, and to innovation. A tall order, I know, to go up against folks with academic credentials and me with just experience and a determination to get things done well. I’ve broken with a large chunk of the technology transfer community by speaking out. I’ve done my best to check my work, to consider alternatives, to give even the little linear model its worthy place as part of the answer. I’m also prepared, if faculty in particular choose administrative domination and orderliness, to move on to other things, let captives be captives and illusions stand for realities, ignorance is bliss and all that. But that is not yet.
A present assignment is an effort to secure transfer of ownership in an invention or work of authorship prior to the invention or work being created. It is distinguished from a promise to assign future inventions or works by the use of assignment language, such as “hereby assign,” which is intended to operate when an invention or work is made that meets the conditions of scope for the assignment. For there to be a present assignment, there has to be ownership of something (now or in the future) and the assignment has to be sufficiently clear that the something is within the scope of the commitment to assign. Thus, know-how isn’t something that’s owned, and therefore cannot be assigned, and therefore cannot be presently assigned. Know-how is practical knowledge. One may direct those that have know-how, and they may agree not to use their know-how except as directed, but there’s nothing to own. One can translate know-how into technical information that is kept secret, and then there is a trade secret, and one can “own” trade secrets, but only in the sense that one possesses the information and others generally do not have it, and reasonable precautions are in place to prevent disclosure. But “own” here is odd. I may be able to give up use of know-how, but not the knowledge of it. Thus, “assignment” cannot mean a transfer in the sense of “giving up” possession, but only of the “enjoyment” or “use” of the knowledge. Let’s keep, then, assignment to statutory stuff, like inventions and works of authorship. And since we are dealing with technology transfer, let’s focus on inventions.
There is a truism that copyright vests automatically and patents have to be applied for. True enough, but what’s not pointed out is that ownership of invention vests automatically as well, as a function of federal law, with the inventors. The challenge is that ownership of invention is something of a prospective right, the right to apply for a patent on the invention, which translates the ownership of invention into a legal right to exclude others from making, using, and selling the invention without permission of the patent owner.
A present assignment aims to move the prospective right from the inventor to a new owner, and to do that for each invention within scope of the assignment, when the invention is made, and without the need for further formalities. Any additional documents saying “assignment” are ceremonial memorializations, as the present assignment actually transacted the business of assignment. [Or if you take Justice Breyer’s argument, present assignment language, like promises to assign, transfers only equitable title, and the later assignment provides legal title]. If the present assignment isn’t expected to transact business this way, then we are back to a promise to assign, badly written.
Most universities have at some point adopted a review + request model for invention ownership. Originally, it was more along the lines of, give us the first chance to request ownership. Then was added a promise to assign those inventions that the university really had to have–particularly by operation of contract–either sponsored research, conditions of donation, or in work commissioned by the university. The driver was sponsored research contracting. If a sponsor required university ownership, then the university really had to have something in place with employees that provided for that ownership. A promise to assign was just the thing. It could be made a condition of participating in sponsored research (as has been the case at MIT) or it could be made a condition of employment (as at the University of California or Stanford).
Federal contracting, with Bayh-Dole, took a different route, however, and did not require university ownership. Bayh-Dole recognized that different universities had different arrangements for invention management. Some managed directly, some through affiliated research foundations, some under contract with organizations such as Research Corporations, and some did not get involved at all, and left things to their faculty inventors. Some universities were voluntary about it, and some were compulsory. There was a diversity, a number of ways of doing things appeared to be productive–Wisconsin was as good at it with WARF and a voluntary approach as MIT was with its compulsory in-house office. Overall, the diversity of approaches appeared to be doing better than the federal agencies were doing–and perhaps were doing differently than the agencies, with more exclusive licenses, less overhead to close deals, and more motivation to find partners that wanted to put things in play. Relatively few universities used an internal model. Most had one or more foundation agents. It is no wonder that in its heart, Bayh-Dole relies on an agent approach to invention management.
Now over thirty years a series of shifts has taken place. From one perspective, they might appear reasonable, even as progress. From another, they will appear as forms of institutional “capture” of government-supplied assets, designed to maximize institutional upsides obtained by exploiting the regulatory framework. Key to all of this, as university research policy responded to Bayh-Dole, has been the misapprehension that Bayh-Dole requires university ownership, and that to get that ownership, folks need to have an IP policy and patent agreements that make assignments a requirement whenever a review of invention circumstances indicated federal funding. It was this little slip, that “elect to retain title” meant “elect title” or “has title” or “is the only entity authorized to have title,” that has shaped the build-out of ownership claims in university policies.
When one looks for the rationale for university ownership claims to inventions, in early formulations, it is to prevent anyone in industry from denying access to an invention for a broader public. Only with MIT in the 1940s do we find a desire to just make money, an attitude that shapes MIT’s practice to the present, and has become a mantra for any number of other universities in making adjustments to their policies. What is good for east-coast private universities must be good for everyone. The primary motivation, however, for university ownership has been compliance with contracts. The argument has been, federal law requires university ownership of inventions, or grants universities the right to take ownership of inventions, or to comply with federal law inventors must assign inventions when requested to do so.
Thus, research policies read: the university has an interest in inventions when contracts require it, and that university employees agree to assign such inventions to the university when requested to do so, so the university can meet its contractual obligations. The new addition was: federally supported research is one such form of extramural contract. This all worked fine in a review + request environment. It was of no matter that federal research, by the time of Bayh-Dole, was swamping out all other sources of research funding at most universities. At the University of Washington, in the early 60s, departments were debating whether to take any federal money at all. Now, the University of Washington brags about its standing as a top recipient of federal funding–typically about 75% of its extramural funding is federal, in comparison with industry funding less than 5%.
Thus, a provision in research policy that a university would take ownership of inventions when contractually required to do so, or when it specially commissioned the work, became the dominant expectation with the passage, and subsequent misinterpretation of Bayh-Dole. A promise to assign did not matter much for Bayh-Dole, if federal law mandated the assignment on request. It was fine to have the promise there, “just in case,” but university administrators were fine with a review plus request format–although for federal funding, it was now interpreted as review + require. Indeed, Bayh-Dole appeared to support just this approach, by requiring inventors to disclose to “personnel identified by the contractor as responsible for patent matters.” That would be technology transfer officers (though Bayh-Dole does not out and say this at all–the personnel responsible for patent matters could be at Research Corporation, or be department heads, or principal investigators).
The shifts, then, from taking an ownership when required to do so, relying on a promise to assign, to believing that federal law provided an ownership position for federally supported inventions regardless of employment obligations, led universities to the idea that federal law was there to allow them to make money by licensing inventions, and that it was only right that all university creative talent submit their inventions, works of authorship, data, software, anything at all that might be worth money in a licensing context, to the university for exploitation. It is in this context that Stanford v Roche comes about, and shows up the problems of interpretation that have been allowed to grow in university administrations.
The Supreme Court in Stanford v Roche ruled that federal law does not vest title with contractors, but merely makes uniform agency behaviors relative to requests by various agents to retain title to federally supported inventions that they have obtained by means of assignment. For a university (or other agent) to get title, there has to be an assignment from the inventor. This really wasn’t shocking news to anyone, since every university obtained assignments of invention as part of the patent application process. No one operated on the argument that Bayh-Dole was a vesting statute, really. But the near variations were sufficient–that federal law required assignment, and limited assignment to the contractor until the contractor had decided whether or not to take assignment.
It all falls apart, however, when it becomes clear that Bayh-Dole does not apply to inventors, but rather to agencies. It does not even apply to universities, particularly. The standard patent rights clause is the means by which agencies engage universities. Thus, the law folks looking in the USC are looking in the wrong place. It is the CFR where the action starts. It is in the CFR that one finds the binding between the law and federal contracting. It is also where the CFR implements the flow down of the patent rights clause for university (and other contractor) employees who might invent. Anyone reciting the USC without looking toward the implementation in particular federal funding agreements is missing the issue entirely–unless the issue is the standing of federal funding agreements, rather than their operation.
Of the elements of the standard patent rights clause, the (f)(2) agreement is perhaps the most important. It flows down federal interests in inventions from the contractor to the employees. It stands in for any patent policy, employment agreement, and assignment obligations. A university worried about compliance with federal funding agreements would do well to implement the (f)(2) agreement as directed. Instead, university administrators are seeking an private means by which to avoid non-compliance. This brings us to present assignments.
The present assignment has the attraction of being self-executing. Thus, as Sean O’Connor suggests in a recent paper, a university could merely require that all inventions made with federal funding are hereby assigned to the university, and then as inventions met that standard, they would be owned by the university. That at least is the theory. Things are different in practice, not the least of which involves the effect of the (f)(2) requirement and problems regarding scope.
The present assignment also has another attraction, in that it appears to harvest everything without the need for a prior review. It changes a review + request culture to one of ownership + limited release. It is a dramatic turn in policy, and yet university administrators at California and Washington are representing it to faculty as “not a change in policy”! How can that be? It would appear to be horribly wrong. But no one is backing down on it yet, so even after reviewing the matter, these folks insist that there’s no change in policy, and therefore no need to discuss it with faculty, or any employees, before imposing it. It is actually quite audacious, or blind. I can’t quite figure which.
The problems in moving from review + request to ownership + release are rather dramatic. In fact, it is not so much that there are “problems” as it is a paradigm shift. If folks want to do it, they are talking a whole nother language that can’t be done with slipping in a “hereby assigns” in a framework that is otherwise review + request.
The review + request approach, backed by a promise to assign where a contract requires university ownership, is entirely well suited to the other practices by which university research operates–faculty-led proposals and faculty-directed and published work, academic freedom, tenure, open laboratories and prohibitions on classified or proprietary work, integration with instruction, the importance of public access and publication, along with safe harbors from regulations such as export control and ITAR for keeping things open. The review + request approach also works perfectly fine with Bayh-Dole, which was *designed for it*. Moving to ownership + release is an option available in Bayh-Dole as well. Small companies might use such an approach. But it was never Bayh-Dole’s sweet spot, and it wasn’t want made the university experience with technology transfer so compelling.
It is possible that one can get some mileage out of ownership + release in a university setting, but it comes at a terrible price, somewhat like 19th century plantations in the American south. Making money, order, and efficiency are not in and of themselves, unqualified public goods, nor are they the institutional virtues that give rise to a research innovation community. Because a given exploit can be done does not mean it must be done, nor in being done in some places does it mean that it will have the same success when every institution is doing it.
If we add present assignments to the remnants of a review + request policy architecture, especially after the Stanford v Roche decision, we readily get what in programming would be a defective architecture. Fix a bug in one place, and a defect will be created in another. The problem is, you don’t know exactly where. One did not “fix” the problem, but rather exchanged a known problem for unknown problems, transmitted through the architecture, to be discovered by users, by practitioners, relying on the system, and finding it fails in new ways, rather than in the old, expectable ways that have been generally documented and avoided. One would be well advised, then, before making changes in an architecture, to ensure that it is sound before “fixing” it.
What happens when a present assignment is added at employment, in place of a promise to assign, in an review+request architecture? First, it changes entirely the purpose of the review. The review is no longer to ascertain the circumstances and establish to the mutual agreement of inventors, investigators, and administrators (if not also sponsors) that a reported invention meets the conditions of assignment. Instead, it is to finger through what one has got ownership of, to decide if it is worthless or valuable. Even in a compulsory review + request architecture, where the institution makes demands based on scope of employment or use of facilities (suspect as these may be), the review is to demonstrate that the conditions are met, prior to the assignment being requested. Present assignment changes all this.
Second, any elements in policy predicated on release of claims or satisfaction of the reporting duty are also transformed or rendered useless. If the university owns outright, then it is not a matter of releasing a prospective interest, waiving an obligation to assign upon request, or confirming that conditions necessitating a request for assignment have not been met. The university has acquired an asset. Regardless of whether the asset is valuable, it has to follow protocols regarding its disposition. In a review architecture, whether an asset is valuable is largely beside the point if the assignment is not required. Good for you, have have done something good there. Run with it if you can. In an ownership architecture, you have it whether you want it or not, valuable or worthless, it is yours to deal with.
The University of Texas policy, in imposing an ownership architecture over the remains of a review + request approach, goes at it this way. The university owns everything. If a worker wants any other disposition, report the asset and the university will confirm its ownership or release (meaning, reassign) the asset in 20 “business” days. But since it already owns everything, what is being offered in exchange for release? How can such releases be handled on an equitable basis, so there is no favoritism, no better deals now than later, here rather than there?
What then is the point of requesting re-assignment? If the asset is worthless, then leave it be. If the asset is valuable, then the University can simply refuse the request. It doesn’t have to give reasons; it has no duty to the inventor to commercialize or license or make publicly available. Even if the University has no plans for the invention, the ownership policy makes the inventor buy the asset back if it’s really that valuable. Otherwise, the deal is, it is better to suppress the invention rights than allow inventors to seek any other agent without paying. Rather than supporting the development and use of university-hosted inventions–a challenging and worthy task–the ownership policy turns the institution into an IP vampire. There is an economy of effort in this: why try to do difficult things when one can simply claim everything and make the inventors, or others, pry the asset back out? A present assignment at employment betrays an institution’s transformation of purpose from service to creative investigators to self-interested antagonist. It is a profound change. That administrators do not admit to seeing it, once again, speaks to an audacity or blindness–neither of which is particularly comforting.
The present assignment does more than change architectures, however. It introduces new problems and liabilities. One set of problems have to do with scope. One of the moves in the shift of university IP policies has been to take a scope drafted for reporting of inventions and make it the scope of compulsory claims of ownership. Rather than casting a wide net of review to ensure identification of those things reasonably requestable by the university, the administrative move is to make the scope of reporting equal to the claim of ownership. If the university has the power to request the reporting of it, then it stands to reason that it also has the power to claim it. This, too, is a substantial change in policy, but it is easily introduced, almost as a sleight of hand, in policies that introduce a compulsory claim on inventions arising from the university and not conveyed into the university by an external sponsor or donor.
Rather than acting as a mediator for an external requirement, the university asserts its own claims, introduces these self-claims into documents with donors and sponsors, and then turns these claims on not only the scope of interests reflected by the outside support, but by making all employment or use of facilities, essentially, commissioned work done for the benefit of the institution, for the benefit, in particular, of the IP licensing program, and that program, with few exceptions, aims to make money from patent positions. It is that simple: the present assignment with the scope of review from the old architecture makes all IP activity for the benefit of institutional money-making. Certainly this is, writ in the abstract, the idea of “profit-making” ventures. However, even in the commercial world, folks find reasons to set limits on what is so claimed, so that these are consistent with the business directions of the company. Here, the business direction is making money on whatever has been claimed. A number of states view this as abuse of the employment relationship. For some reason, university administrators appear to have no such qualms, even in states with such laws.
The undefined, general scope has another problem. Policies citing a scope of employment for faculty have a significant problem in establishing just what that scope of employment is. Other than assigned teaching and grading and office hours, perhaps some committee and advising responsibilities, most other faculty work is not assigned and directed. Certainly not research, whether sponsored or “departmental.” Faculty ask to be released from their scope of employment to undertake sponsored research. There’s a separate budget under which they get paid, the terms of the sponsored research contract trump policy, and the activity is based on the special performance of the faculty investigators. The issue is not that the university can extend its scope of employment to include whatever gets done in the sponsored project–it is, rather, that what gets done in the sponsored project is up to the faculty investigators.
This is even true for federally supported projects. The faculty generated proposal for funding becomes the statement of work. Under Bayh-Dole, only those inventions that are within the “planned and committed” activities of the grant are subject to the standard patent rights clause. Bayh-Dole anticipates that a contractor might hold back and claim *fewer* inventions than the agency believes are within scope. In practice, however, university administrators have done just the opposite: they have used Bayh-Dole to claim everything they can as federally supported, believing that by making the claim, they also then could force the assignment of title, if they didn’t hold it outright when the notified the agency of their “election of title” (not “election to retain title”). Thus, the challenge for a present assignment is based on what scope is going to control the obligation, and who is entitled to make that choice. If the faculty member intends certain work under a grant, and does other work outside the grant, and outside university resources and time, then it is clearly the faculty member’s decision that determines scope.
Present assignment advocates therefore have to find a way to constrain faculty choices about where to do work. The policies of choice are those involving conflict of interest and commitment. If one invents in a manner that would deny the institution access to IP positions, then the inventor is circumventing institutional benefit for personal gain, and that’s a conflict of interest. Thus, a present assignment with a general scope has to be somehow expanded so that it includes what otherwise would be done informally or in private consulting.
The attempt results in some ludicrous outcomes. Take the University of Texas policy. It says that the university owns all things made within the scope of employment. It also allows that it does not claim ownership of scholarship. But then Texas demands that its facilities and funding only be used for the creation of intellectual property that it owns. Thus, scholarship is either not within the scope of employment or cannot be done using Texas facilities or funding. Scholarship is, in essence, a misuse of university facilities. It is a defect in the architecture introduced by “fixing” the promise to assign policy with an ownership + release policy.
If the scope of employment (or work subject to the present assignment) is largely chosen by the faculty member, then a present assignment amounts to little more than assigning what one chooses to assign, with the removal of a paperwork step that would give definition to the asset. Intention to be bound, is after all, a hallmark of enforceable contracts. On the other hand, if the scope of the assignment is a matter of interpretation for administrators and not within a written contract with a sponsor (which also may have something to say one way or another), then the present assignment is an agreement to agree with a later decision by administrators about scope. In a company setting, present assignments are typically bound up with other employment covenants, such as non-disclosure, non-use, and non-compete with regard to company assets. An invention, before it a thing owned under patent laws, is claimed by a company as a trade secret. The information an inventor is working with is claimed by the company. The facilities are provided for use to benefit the company. The scope is established through multiple elements, not simply on a decision as to what a faculty member’s scope of employment is, at some future date, based on whether it sounds like something the university should have.
Without a defined scope established by the university based on its business–which for state universities is a matter of charters and statutes, not the whims of administrators–a present assignment is potentially worthless because it lacks sufficient scope to establish there was a mutual intent to obligate a particular invention, since even the administrators cannot say what they intended–except for anything–and they cannot establish that this was the intent of the faculty inventor in agreeing to the present assignment. The present assignment acts much more like an administrative wish list–that faculty will accept their wishes when they make them known, no matter the defects in the drafting, lack of definition in the scope, or failure to have a meeting of the minds with regard to interpretation. As long as faculty choose what the work on, and where they work on it, and for the benefit of whom they choose, scope will be the problem in compulsory assignment architectures, not the manner in which the assignment is executed.
The present assignment also creates liability for the university. As a portfolio grows, a licensing program faces multiple challenges that do not show up if one considers each invention in isolation (which is however how most university policies are drafted, as if each invention takes place without context, with one inventor producing one invention, and lacking idea what to do with it, assigns it to the university, which produces one license, which results in royalties–everything else is just doing this repeatedly). These issues include background rights, improvements, implied licenses, and double licensing. We will deal briefly with each, enough to show the impact brought by a present assignment move to an ownership + release architecture.
Background rights are rights held by the licensor (or known to the licensor) that are required for the licensee to practice the invention in the manner described. A licensee may obtain a license to an invention, but if the licensor holds another patent (say more general than the one licensed), then the licensee still does not have sufficient rights to practice. Similarly if the product envisioned requires two steps, but the licensor has only revealed it has patent claims on one of the steps. As a portfolio increases in scale, the review of inventions and other assets (software, for instance) that may form background rights becomes much more challenging. Disclaiming background rights is not a help, because doing so sends up red flags for the licensee that they may be hung out to dry if they pay for a license and find they have not got all the licenses they need from the university.
A present assignment approach vastly expands the university ownership position without notice. One will have to disclaim background rights, because there will be no way to know what is owned until it is reported, but it will be obligated otherwise without notice, if it is not disclaimed. At issue is also a second licensing problem, that of implied licensing. An implied license is one that is granted without an express grant of rights–through actions, or by apparent necessity. If I have a piece of software, and I say, sure you can use it, then if I also have a patent on the software, I’ve granted an implied license in that patent. Imagine a university with hundreds of labs. One develops an algorithm and the university claims it as an invention. Another lab develops a piece of software and makes it available to a user community under an open source license, like Apache License version 2, which also commits any patents owned by the party providing the code. Well, if the present assignment reaches to this software, then the university owns it, and has granted an implied license to the algorithm invention in permitting the release of the software under the open source license.
Take it one step further. If the university licenses the algorithm patent exclusively while there’s an open source release under an implied license, then there is a double licensing problem for the university. Double licensing means granting a sole or exclusive right to more than one licensee. Something will have to break. An easy way for double licensing to occur is to have two laboratories working in related areas, one with a sponsor having a first right to negotiate an exclusive license, and the other working with a sponsor who has a royalty-free non-exclusive license. If the two laboratories co-invent, and the university owns both inventions, then there’s a double licensing problem, as the university cannot honor the first right to negotiate an exclusive license and grant a non-exclusive license to another entity. If the university does not own either invention until it requests ownership, it has not created the the liability. The more expansive and less documented a portfolio, and the more certain a licensing organization’s claims to it, the more problems it will inevitably have with double licensing. If the institution disclaims double licensing, it may as well refuse to license anything. The present assignment is a little engine that greatly expands a university’s undocumented portfolio of owned assets. Every license has to be responsive to those assets, even if no licensing officer is aware of them. One will have to take a conservative position–a hopelessly conservative position that aims to transfer all the risk onto the licensee.
Finally, we have improvements. When a license is granted, and the licensor is involved in active research and development, the licensee reasonably wants improvements. These can be of various sorts. Improvements can mean variations within the same patent family–divisional patents in the case of multiple related inventions, or continuations in part that add new subject matter but claim priority from the original patent application. Improvements also can mean new best modes of practicing the invention, or even new methods that work around the claims of one invention to do something functionally similar, or even better. In engineering, this sort of successive improvement is common. One can go through 30 variations in a year, each providing some advantage or expansion or change in how a particular problem or property is dealt with.
As with background rights, improvements can create challenges for a university running multiple laboratories in related areas of work. If a licensing office includes improvements in a license, which is a reasonable thing, then the scope of those improvements may include university owned assets that have not been disclosed to the licensing office, or worse, have been disclosed but not filed in a way that the relationship between one technology and another is obvious. If the scope for improvements is narrowly drawn, then it is easy for the university to authorize work-arounds that frustrate the advantages of an exclusive license. If the scope for improvements is broadly drawn, then the university risks implied licenses and double licensing that can devastate a licensing program. Owning too much is a licensor’s nightmare. This nightmare is never addressed in reports of invention disclosure activity. A technology transfer office never reports it has had enough and is working to limit the number of assets it owns. But offices should do this. A present assignment architecture makes it impossible to stop, even if an office does not want to own, cannot afford to own, an invention or other asset that would put in jeopardy the license agreements it has made in sponsored research and in its own licensing and startup programs.
The point is, a present assignment strategy, which may be well suited to a company environment, with a controlled workplace and limitations on the public release of information, software, and publications, is a nightmarish strategy for a university licensing program, which must deal with open labs, faculty publications, student and visitor volunteers, duplicate efforts, and a range of relationships and commitments to sponsors. There is a simple choice for the licensing program: limit what one owns and is responsible for, or try to shut down the openness in favor of corporate controls. A present assignment, as being advocated by legal folks everywhere, is entirely the wrong thing for university licensing programs. The problems of background rights, improvements, implied licenses, and improvements will force licensing programs to be conservative, highly selective, and unwilling to make low-value assets broadly available for fear of breaching a high-value deal or precluding one from happening later. In all, present assignments with a broad scope, as a self-executing engine to collect everything it can, early and often, is exactly not what a university licensing shop needs.
This brings us, finally, to two remaining matters involving present licenses. The first of these is the problem of when a present assignment operates. The second has to do with whether the policy that demands a present assignment itself is defective and unenforceable as a result of the improper introduction or operation of the present assignment. We will look briefly at each.
When does a present assignment to an invention operate? An invention is made when it is conceived and reduced to practice. These are terms of patent law. If we are dealing with any broader definition of invention, then we are not dealing in patent law, but in something else, such as trade secret. How a university can claim ownership under trade secret is something else entirely–companies do it all the time, but then, they are fundamentally different in their policy architecture and terms of employment. Conception takes place when all the elements of an invention are present in the mind of the inventor, and the inventor recognizes the elements as an invention. Reduction to practice takes place either by producing a prototype that exhibits all of the elements of the conception and performs its intended function (actual reduction to practice) or by producing a patent application that describes the invention so that one with ordinary skill in the art could practice the invention without undue experimentation (constructive reduction to practice). One might think that an invention comes into existence with conception, but no–reduction to practice is also required. Then the invention is made, and then a present assignment directed at inventions operates. The assignment does not get a “priority” date of when the obligation to assign was created. Rather, it happens when the invention is “made.”
Now look at the Stanford v Roche situation. Stanford had a promise to assign at employment. Then it permitted consulting with an obligation to assign (in the form of a present assignment, but it doesn’t really matter). Now Stanford allows the employee to join research at the university in the same area as the consulting. It cannot re-apply its promise to assign, because it the consulting was out of scope, and then approved the employee’s participation in the research, knowing the consulting obligation. But the university really wants ownership. It wants to be able to require its employee to assign to the university even though it knows of the assignment obligation to the company. It cannot navigate 35 USC 261. It is hosed. How would a present assignment at employment change the outcome? The consulting is still out of scope. The return is still allowed. An invention gets made. Does the present assignment control because it is *earlier*? No. The later assignment controls.
The way for the university to get at this is to 1) not know about potentially competing assignments and 2) obtain a later assignment and record it first in the Patent Office. For this, an excellent practice would be to place the present assignment (or assignment obligation) at the point that individuals join a project in which IP is going to be claimed. Then the present assignment is the most recent. If an individual cannot sign because of prior commitments, that’s when it comes out and the prior commitment is renegotiated or the individual is excluded from participating. If at the same time an agreement is made that there will be no commitment of inventions that would come within the scope of the controlled IP project, then one has a reasonable set of agreements to manage the project and expect ownership. If one makes commitments to the sponsor only to this level of management, one has a decent change of navigating ownership. Present assignment at employment, however, won’t do it so long as consulting and publication are open. If one closes off consulting, then one in effect prevents faculty from making some of their most important contributions to industry.
Another route that universities have used to manage this situation is to require employees who consult to add a line in any consulting agreement committing IP that obligations to the university take precedence. This allows the consultant with conflicting obligations regarding assignment to choose the university, and thus avoid getting sued by someone. However, a present assignment at employment doesn’t then change the outcome. A promise to assign, even if it is at most equitable title, is just as good as a present assignment–if that obligation has precedence, then any conflicting obligation cannot be intended to operate in the consulting deal. The present assignment then adds nothing.
Of course, a company knowing these matters may well change its practices. One strategy would be to limit the consultant not to join discretionary projects that would introduce obligations in conflict with assignment obligations in consulting. That is, don’t use university resources or undertake self-chosen scholarship or join extramural research that would be in the same area as assignment obligations to the company. The university would then have to direct the employee to undertake the work in order for its ownership policy to have any effect. A second strategy is for the company to notify the university with regard to any commitment to assign made in consulting. Then 35 USC 261 kicks in, and the university has a problem with asserting a later claim of ownership, regardless of whether it started with a promise to assign or present assignment.
In short, the present assignment solves no ownership problems, introduces new intractable ones, and has direct workarounds for companies. Existing strategies, such as precedent language and using agreements at the time of entry to controlled-IP projects (such as implementing (f)(2) agreements for federally supported work) do the work required as well or better than an ownership + release approach.
Assumptions about university patent portfolios make present assignments more attractive than they should be. For the attorney, it is all about acquiring as much property as possible, with the assumption that the portfolio consists of unrelated, independent inventions, that there will not be any overlap in scope of committed rights, and that licensing officers can maintain a comprehensive list of all properties that are owned, along with their attributes, possible applications, lines of development and improvement, and ancillary related assets such as data, software, and the like. This is a simplifying assumption that breaks down almost immediately in practice, where information is incomplete, technology changing, multiple arrangements happening across multiple offices simultaneously, and a host of undocumented technologies and relationships color transactions. Add in badly written policies and one has all the makings of problems. All the present assignment does is increase the intake of, and responsibility for, undocumented assets. It is a net liability.
There is an argument that the present assignment does not operate until a work has been reviewed after disclosure. But if that is the case, then it is not a present assignment that one wants, but a conditional assignment, based on the results of the review. That would be a present assignment for a later date: “If the university determines after review that an invention I have made is within the scope of its ownership claims, then as of the date of that determination, I hereby assign my entire right, title, and interest in and to said invention to the University. I reserve the right to appeal the university’s determination of ownership.” Or something like that.
Such a conditional assignment allows for review, but then it does not do much that’s different from the old review + request approach, other than make the assignment happen as of the date of the determination, rather than when an assignment document is drawn up and signed by the inventor. But this approach doesn’t help anything, as the problem isn’t just getting ownership expeditiously, it is overcoming any intervening conflicting obligation to assign. Such conflicting obligations do not arise because of a failure in a policy claiming ownership, nor in how that ownership is acquired, whether by a promise to assign or present assignment. The “problem” if it is such, is allowing anyone to have any assignment obligations and then later act to come within the scope of university claims of ownership.
The university could shut down consulting that involved assignment commitments on any subject matter. But that would not cover the cases in which faculty or staff or students are hired with such commitments already in place. Some corporations send their employees to graduate school and do not release them from their IP obligations to the company. If the university insists that all use of facilities gives rise to an institutional ownership claim, even for students, then the institution’s policy really is: no student can receive an education that involves use of facilities if they have made IP commitments to their employer. And this is all to pack as much as possible into a technology transfer program, to its disadvantage and increased liability. Shutting down consulting and prohibiting prior and continuing commitments is not particularly realistic. It will be impossible to enforce, it runs against public policy, it turns the university into a fortress operated for the tech transfer office, and it will drive away creative people and company partnerships. It is almost that this is, indeed, what senior university leadership wants to have happen. I have heard leaders say, when it comes to industry sponsored research, that it would be better if it all went away.
If a policy is hopelessly not enforced, with a scope of claims that cannot be determined without review, and without perhaps also a statement of intent on the part of the employee obligated to the policy, then what is the status of that policy as part of the employment agreement? Consider, if a present assignment cannot be determined to operate without a review, does it operate anyway, just banging away without any contribution of intent by someone who as been told that as a matter of policy they have agreed to present assignments? An assignment takes place because there is an act of intent to assign. The paper records that intent, but it does not construct the intent when there is none. If a policy is not enforced, if it cannot be understood to operate, if there is no indication that there is a meeting of minds, then it would appear that the policy is not acting as part of a contract, but as part of a set of laws that can be changed and interpreted at will by university administrators, regardless of the understanding or intent or reasonable interpretation that any employee (or student or visitor or collaborator) might give to the words of the policy. Such a thing is not a contract, because it does not meet the conditions for a contract. And universities, even public ones, do not have the power to make laws in place of employment agreements or IP agreements. Confusing contractual obligations and administrative rules with laws got university administrators into trouble with Bayh-Dole. It would appear that this confusion extended well beyond Bayh-Dole into IP policies and the basis for agreements with faculty and other university personnel.
The architecture that has been created by this conflation of obligations is horribly defective. Present assignment practice is ill suited to it, and will create problems elsewhere in policy and practice. Present assignment does not address the Stanford v Roche situation of conflicting later assignments. Other practices do and would address these situations. Claiming a present assignment will do the job when it won’t not only creates problems, then, but also may create a false sense of security, when there is no replacement for vigilance and selectivity, and certainly not continued reliance on a policy architecture that is demonstrably defective.