Here’s more evidence in the wild about how administrators are warping the Stanford v. Roche decision. Here it is the University of Washington, sending out a note to faculty about little technical changes in approval forms for consulting, claiming it’s not a change in policy:
The changes do not reflect a change in policy, or any increase in intellectual property rights to the University. Rather they are to provide guidance and to ensure the rights of the University are not lost due to us not using current standard assignment language that other Universities and our industry partners have adopted.
The new language to follow is the “current standard language”. More than that, “industry partners” have adopted it. This is a strange opening, even without the spurious capitalization of “Universities”. Why would company present assignment language have any bearing on university patent practice? The cover note goes on:
The form has also been modified to follow guidance following the recent Stanford v Roche holding by the Supreme Court. This case highlighted that the present assignment language in consulting agreements could supersede under certain circumstances the prospective assignment language of University employment documents and policy. The new language in the outside work approval form will help prevent the uncertainty concerning intellectual property that gave rise to Stanford v Roche, and protect the University of Washington, and the UW inventors who share in proceeds from the intellectual property under University of Washington policy. UW Human Resources is also changing the outside work approval form for professional and classified staff to reflect this guidance.
Stanford v Roche at the Supreme Court had nothing to do with present assignment language. It had to do with Stanford’s claim that it didn’t need an assignment at all because federal law vested patent rights with the university outright when it received federal funding. The Supreme Court flat rejected that argument. No surprise there, given the argument had no support in Bayh-Dole, patent law, or public policy. But then even university administrators can pine for a totalitarian state.
The present assignment language was involved in the CAFC ruling. There it was merely the means by which a valid assignment was made in a consulting arrangement encouraged by the university. It did not trump a prior promise to assign, because the university in effect waived that promise when it allowed the consulting. The CAFC’s issue was whether the valid assignment to the company was voided by a later attempt by Stanford to obtain an assignment for the same invention. The CAFC said it wasn’t.
[clarification and further comment 8/27/16: the CAFC also argued that the Stanford promise to assign did not operate until there was a formal assignment, while the present assignment of the Cetus agreement operated without the need for a further assignment. It’s actually even squeaker than this, because the CAFC only got so far as to establish the effect of the present assignment was that Stanford could not establish standing to sue Roche, not that Roche ended up with legal title to the invention. So it is not at all clear that the CAFC decided that a present assignment was the same as assignment of legal title.
Had there been two competing promises to assign, the court would have had to look further to find a basis for differentiating the two assignment promises. The Stanford promise to assign was out of scope of Stanford interest until Stanford permitted its employee to return from work at Cetus and join a federally funded project working in the same area. At that point, the fact that the CAFC found that Stanford had notice of the promise to assign to Cetus would no doubt have become significant. Stanford knew that its employee had promised to assign to Cetus. Stanford approved of the arrangement, even if its licensing office didn’t know about it. And yet Stanford assigned the employee to work at Stanford in a project that Stanford claimed had obligations that conflicted with what Stanford had already approved. Bayh-Dole was asserted as an exploit that permitted Stanford to void its employee’s promise to assign to Cetus. The court may well then have taken up Bayh-Dole and found that it does not require Stanford to own inventions and therefore the employee’s initial promise to assign to Stanford was simply not within scope. Then we would be left with a Stanford-approved promise to assign to Cetus and seven years later, without notice to Cetus of an invention or respect for the promise Stanford had approved as a condition that its employee gain privileged access to Cetus research and technology, an assignment to Stanford of rights previously obligated.
Is this a matter of justice approving clever university maneuvering? Or would the CAFC have looked to find some way to require Stanford to honor the promise to assign made by its employee and approved by Stanford (even if not by its licensing office)? I could imagine the CAFC finding that Stanford misrepresented Bayh-Dole to its inventors in seeking their assignment, had interfered in a business relationship between its employee and Cetus, and that even a promise to assign resulted in the encumbering patent rights such that even if no assignment document ever transferred legal title to Cetus, the inventor was no longer free to assign those same rights to Stanford.
The promise to assign to Stanford was out of scope at the time the promise to assign to Cetus was made. There would still be no conflict. The promise to assign to Cetus remained valid when the invention was made at Stanford. Bayh-Dole did not require Stanford to own–it merely (through the standard patent rights clause) gave Stanford a right to keep title if the government requested title. Stanford’s own policy was one allowed inventors to own whenever possible, but that either Stanford or a sponsor or research would own if a research contract so required. Since Stanford is not required to own, but merely has a basis to resist a federal request for ownership, Stanford in claiming it must own the invention would have distorted its own policy into one that forbade its inventors from assigning to the government as a sponsor of research–despite Stanford’s own patent agreement providing for just such an outcome.
I could imagine, then, the CAFC expanding its look at contracts. It would not have had to find that Bayh-Dole did not vest title with Stanford–that still is a horrible thought (the Supreme Court called it “deeply troubling”). All the CAFC would have had to find is that Stanford had not obtained sufficient rights from Bayh-Dole to trigger the original promise to assign, and that its inventor did not have sufficient rights to assign when Stanford did obtain a written assignment that Stanford obtained the standing necessary to sue for infringement. For that, perhaps it would have considered 35 USC 261 in a different light:
An interest that constitutes an assignment, grant or conveyance shall be void as against any subsequent purchaser or mortgagee for a valuable consideration, without notice, unless it is recorded in the Patent and Trademark Office within three months from its date or prior to the date of such subsequent purchase or mortgage.
Perhaps a promise to assign is an “interest that constitutes a grant or conveyance.” Or perhaps a promise to assign is an assignment, not of legal title, but of equitable title. But promises to assign are not usually recorded in the PTO, so one might argue that to secure rights, all promises to assign–and by extension all present assignments of future interests–must be recorded in the PTO to be effective against a later assignment. The problem in all of this is the bit about “without notice.” It’s clear that an assignment timely recorded in the PTO defeats a later assignment. It’s clear that an assignment made for “a valuable consideration” and made without notice of a prior assignment that has not been recorded defeats that prior assignment. But there are plenty of gaps. What if the later assignment is with notice but the prior assignment has not been recorded? What if the later assignment does not involve “a valuable consideration” but is made by assertion that federal law requires it? What if the later assignment also is not recorded in the PTO and the prior assignment then is recorded?
Given the circumstances of the case, Stanford would have had no claim to the invention even with a present assignment in its patent agreement, even with a claim to own anything developed at the university, whether sponsored or not, required to be assigned or not. The invention was out of scope of such a policy and its apparatus. Bayh-Dole could not save Stanford’s claim. The situation comes down to one in which Stanford ignores a prior obligation with regard to an invention–an obligation that it has approved and which is entirely well formed in policy–and takes the invention for itself, and then turns on the party to which the invention had been obligated when that party won’t pay. What’s just?
Of course, Stanford folks must have rationalized all this–the invention wasn’t obligated, really, it was different somehow. The invention did not result from access to confidential information (even though the agreement signed with Cetus concerned only whether the invention involved the business interests of Cetus, then or in the future). Federal law gave Stanford the right to take title (though it doesn’t). The company didn’t record its interest or follow up when it first found out that Stanford had got a patent. (Of course, it doesn’t appear that Stanford asked about the agreement with Cetus or notified Cetus that there was an invention, and the like). And they have to defend their patents or risk in the future people brushing them off, and they have to defend their patents when they can make money, because they have promised a share to their inventors (so the litigation is meant to teach everyone else a lesson that Stanford can come after them, too). Or whatever the rationalizations might be.
Sometimes, I think, justices reason from what is just to a line of precedents and findings that support that sense. The particular argument, unless it is later overturned, illuminates some bit of the law and informs subsequent practice. But that particular argument does not necessary go off and illuminate the law in its entirety, examine a bunch of alternative scenarios, and provide a handbook for the law forever after. The CAFC sought to prevent an outcome in which a federal law allowed federal contractors to invalidate contracts between other parties without notice or due process. That concern, too, was a key point in the Supreme Court’s finding that Bayh-Dole did not and could not vest title to subject inventions in federal contractors. To do so would be to permit those contractors to void private contracts made by those it assigned to work on federally funded projects. Now that would be title uncertainty writ large and with impunity.
Lapsing into hypotheticals does not get us to practice certainty. The law has plenty of practice issues left unaddressed by the courts. But I can imagine the CAFC doing the right thing, using different arguments than it did use. Especially if Bayh-Dole had been properly presented from the outset. Perhaps you can imagine such a thing, too–or perhaps you prefer to imagine the CAFC endorsing Stanford for regretting (or forgetting, or dismissing) the arrangements that allowed it to make the invention in the first place and then coming back to hit up the party it took the rights from for a quarter billion dollars. That’s clever, but is it just? Are federally supported inventions so very important that we will sacrifice the freedom to contract for them? Is making money by asserting patent rights in inventions that have made it to actual, beneficial use more important than that actual, beneficial use? What society do you choose to hope for?]
Even if Stanford’s promise to assign language had been a present assignment of expectant interest, it would not have operated because Stanford had waived that expectant interest in approving the consulting deal. It had no expectant interest in what its policy said wasn’t within its scope of claims. Later federal funding doesn’t change that. You can’t void a deal–the approval for consulting–unilaterally just because you get money from somewhere to work with what you know to be committed elsewhere.
[Update. Another UW communication to faculty, this time from the Executive Vice President, makes it really clear that the administration wants to make people believe that somehow Stanford v. Roche means “claim more IP, claim it expediently, eliminate the foolishness in policy about there being options or flexibility:
“Please note that the approval request form has been modified to follow recent guidance from the U.S. Supreme Court; you must use the new version of the form, dated October 2011. Questions regarding the revised language should be directed to C4C at: UWC4C@uw.edu or the IP Hotline: 206.616.8776.”]
Here is the bit of the UW form that has been changed (my bold for comments to follow–the change I am focused on is the “and I hereby assign” bit):
The Statements above are truthful to the best of my knowledge. With this request for outside work, I acknowledge that I am bound by and I agree to comply with the University Patent, Invention, and Copyright Policy (Executive Order 36) (“Policy”), as it may be amended from time to time. In accordance with this Policy, I will disclose all inventions and discoveries I create to the UW Center for Commercialization, including any that I create in connection with any outside work. I agree to assign and I hereby assign to the University all my rights in any intellectual property to which the University has a right of assignment under the Policy, provided I created such intellectual property in the course of my University activities or responsibilities or with more than incidental use of University resources.
The text here is bizarre. Let’s survey this little haunted house. I’ll only touch on the problems. This is a form requesting approval for outside consulting, not an employment agreement or patent policy. First, what is the purpose of the “agree to be bound by” language? If it pertains to employment, then that has to be covered elsewhere–that’s how the policy is bound into the deal in the first place. If it is aiming to extend to the requested consulting, then that’s a huge extension, because the consulting is out of scope for employment. So what is the employer doing here? Seems potentially unethical. It’s like throwing in a little language that doesn’t appear to matter but later will turn out to be material–that the consulting is now covered by the policy claims.
Second, it extends the agreement to amendments of the patent policy–without notice, without factoring in the existing consulting approval. That’s just not right. How could any company working with a university consultant rely on this? The university could change the policy, void any deal, and claim that the faculty member had agreed to the changes. That’s also not ethical. And it’s lousy public policy. Perhaps the university should come out and say, “We discourage any sort of consulting that would result in anything valuable being produced by faculty.” Talk about an ivory prison.
Next, the form text requires disclosure of inventions and discoveries without regard to whether they are patentable, and without any obligation by UW to hold such disclosures in confidence and not use the information disclosed in any way. Note that disclosing an invention per the UW policy has next to nothing to do with making a determination of whether the university has any claim on the invention. For that, one has to examine the scope of research agreements to which the consulting employee is party, use of resources, and whatever it is that anyone can construe as the University’s scope of claim. The particular characteristics of a given invention or discovery don’t advance such a determination. All that is necessary is to report the fact of a potentially patentable invention, the general area of its subject matter, with a statement that no university resources were used. If the claim is not true, then it’s a problem for later, when a patent application is published, for folks to thrash it out.
Furthermore, by agreeing to disclose the invention to UW without any confidentiality, it would appear that this will count as a public disclosure and can potentially blow patent rights for the company. If I were in a company, I would not hire any UW personnel as consultants without this language being struck. Its only role is to create uncertainty and set up a way for UW to try to void an approved consulting relationship whenever it wants, by changing policy, demanding disclosure, and refusing confidentiality.
Now we get the “hereby assign” addition. This is a huge change. UW’s policy is that it claims the right to obtain ownership of inventions, but that it also may waive that claim.
For an example of a really poorly drafted policy, read it here.
(For fun–note that it applies to as well to “nonpatented innovations”–a rather recent addition to the policy reflecting deep misunderstandings of what a patent and invention policy is all about. “Patentable” and “invention” are the watchwords. An “innovation” is something entirely different!)
My gosh, and it doesn’t get any better. Employees are required to report inventions, but agree to assign only those inventions “in which the University has an interest” (Section C. if you are following along at home) and even then, it could be to the University, to an invention management agency, or to a sponsor of research. But here, in this newly revised form, it’s a present assignment to the University. There’s no option to assign to a management agency or a sponsor. It is like these parts of policy have been lined out. That’s quite a change in policy, protestations to the contrary not withstanding. Further, if the University were to agree that inventions were to be assigned directly by inventors to the sponsor (as is the case, potentially, with Bayh-Dole!), then the consulting form language would breach that deal. What a fustercluck.
No, we’re not done. The present assignment further applies to all intellectual property–not just to patents. What does that mean? Are we talking statutory IP here–patents, copyrights, and trademarks–or is this the sloppy expansive version that includes trade secrets, domain names, data, and whatever ideas I may have in my head? No one knows, because no one bothers to define terms here. We will leave pass the “all my rights” as distinct from “title”. In UW patent policy we read, “Inventions in which the University has an interest but which do not meet University criteria for patenting shall be managed in accordance with policies and procedures determined by the University Office of Intellectual Property and Technology Transfer.” (Section C, con’t). The procedures include “under very rare circumstances, 3) The transfer, for appropriate consideration, of the patent rights to the inventors”. This is totally messed up. First the policy states that assignment will be made only when the University has an interest, then it appears that the University could have an interest, but not in obtaining patent protection (that is, the theory of IP by which an invention is owned), but then even in refusing to file a patent application, demanding “appropriate consideration” to transfer back the “patent rights”–which it would not have anyway, since it had not even applied for a patent.
But now back to this new form, with the present assignment, which isn’t in the UW patent policy. The cover note says this is not a change in policy, but clearly it is. Now, if there’s any consulting, an employee is locked into what is signed on the request form, not what is otherwise in policy. That’s a huge change. If the request form merely recited the policy, fine. No need for a new signature. But if the request is different, and there’s a signature, it looks like a new agreement. If it is, it’s a change in policy, or worse, it is a claim without any of the formalities of policy–like review and discussion. And if it’s not an agreement, then it’s not binding, and it’s a waste of ink to have put it in.
If one consults, then apparently one assigns all IP to UW, even though UW’s own policy does not require assignment outright but makes assignment conditioned on a formal assignment. There are further differences between the form and the scope of claim in policy. The scope of claim in policy is pretty wretched stuff. I won’t go into it here. It’s enough that folks can’t recite the policy properly and instead limit the interest in consulting work to stuff created in “University activities and responsibilities” (but that has nothing to do with consulting, which is a request for something *other* than university activities) or with “more than incidental use” of University “resources” (so, take the consulting off campus). That is a huge restriction on the scope of claim set out in the patent policy, which includes inventions “related a) Directly to the business of the University, or b) The University’s actual or demonstrably anticipated research or development”. That language is drawn from RCW labor law, where it is intended as a limitation on claims by employers on employee inventions, but here introduced to *expand* the university’s scope of claim. So which statement of policy controls? It’s hard to say. It appears that the folks drafting this really don’t care.
In any event, this last construction regarding assignment has nothing to do with the consulting request. Here it is again:
I agree to assign and I hereby assign to the University all my rights in any intellectual property to which the University has a right of assignment under the Policy, provided I created such intellectual property in the course of my University activities or responsibilities or with more than incidental use of University resources.
It does not say anything about consulting. It is clearly a change in policy, introduced opportunistically into a request for consulting without review, discussion, or formal processes of approval. That’s all there is to it. How does a university get away with such slop passing for policy? Is it gross incompetence? Bullying? Love of power, regardless of the reasoning? I just don’t get it. How can folks look at this stuff and not think just how clueless the UW Center for Commercialization is.
The introduction of present assignment language in university agreements short-circuits options in policy statements that the administration may not claim all inventions. Instead, all inventions that can be claimed are assigned, and the business then is what to do with that ownership. The effect on consulting is not to protect the university and its inventors, but rather to make a claim to take as much as possible, and doing this merely increases the uncertainty as to the scope of claims, adds bureaucracy to private agreements, and in the case of public universities, uses (abuses) state conflict of interest laws to impose claims on personal property (inventions) that universities (including University of Washington) were unable to get by invoking (badly, wrongly, foolishly) federal law to do it.
There was no uncertainty in the consulting in Stanford v. Roche. The consultant’s boss was on the technical advisory board of the company. It was by means of the consulting that his post doc was able to learn the technology (over the course of 9 months). It was the capability with the technology that got the federal grant and directly gave rise to the invention (or so the courts held). The only thing that’s uncertain is why Stanford tried to break the deal that had been made.
But now we have this “Stanford legacy”–or maybe we should call it the bozonet legacy, given how it has spread–in which university patent policies are being altered to a corporate standard, on faulty premises, often with unbearably sloppy implementation, for no good university, research, science, innovation or economic development reason whatsoever.