Present Assignment of Future Invention Rights: Some Heretical Thoughts on the Stanford Case*

In my earlier article addressing the Board of Trustees of the Leland Stanford Junior University v. Roche Molecular Systems, Inc., I suggested the Federal Circuit “still doesn’t get it” on how title to “subject inventions” works under Bayh-Dole.  See CAFC Continues to Struggle with How Title to Subject Inventions Works under Bayh-Dole.  My original view was that the title to the “subject inventions” (i.e., those resulting from federally sponsored research) initially resided with the organization (e.g., Stanford University) which carried out the sponsored research. I now confess that I, like Stanford University, overstated how Bayh-Dole works with regard to title to “subject inventions.”  In affirming the Federal Circuit, a majority of the Supreme Court (6 Justices) later ruled that Bayh-Dole “does not automatically vest title to federally funded inventions in federal contractors or authorize contractors to unilaterally take title to such inventions.”  See head notes to the Supreme Court’s slip opinion.

But having discussed in the patent “blogosphere” what the Supreme Court’s decision in the Stanford case says and actually holds, I (like Stanford University) am not alone in “overstatements.”  One of the critical issues in the Stanford case that is glossed over (or at least not addressed directly) by the Supreme Court majority (as well as others in the patent “blogosphere”) is what happens when you have a present assignment (or at least a contractual obligation to assign) of invention rights that don’t exist at the time of the assignment (aka “future invention rights”).  Should (as the Federal Circuit held) Roche (or more appropriately its predecessor, Cetus) by using the language “I will assign and do hereby assign” (aka the “Cetus Assignment Clause”) trump what may have been an earlier obligation  by a Stanford University researcher (Mark Holodniy) to assign invention rights to Stanford University (aka the “Stanford University Assignment Obligation”)?  I would argue, as did Justice Sotomayor’s concurring opinion and Justice Breyer’s dissenting opinion (joined by Justice Ginsburg) that a “yes” answer to that question defies logic, reason, and prior case precedent (other than the Federal Circuit’s 1991 case of FilmTec Corp. v. Allied Signal, Inc. whose logic, reasoning, and adherence to prior case precedent was challenged by both Justice Sotomayor’s concurring opinion, as well as Justice Breyer’s dissenting opinion).

What plagues the Supreme Court’s decision in the Stanford case (and which Justice Breyer aptly notes in his dissenting opinion) is what, exactly, is the correct factual record? In particular:  (1) in what chronological order did the Stanford University Assignment Obligation and the Cetus Assignment Clause occur?; (2) when did Stanford University get the federal funding for the research that Holodniy allegedly did for them, versus when were Holodniy’s research activities carried out at Cetus after he signed Cetus’ “Visitor’s Confidentiality Agreement” (“VCA”) with the Cetus Assignment Clause?; and (3) relative to the invention rights at issue, how much were those rights based on Stanford University’s federally funded research, versus how much were based on the research activities carried out at Cetus?  The factual record discussed in the Supreme Court’s opinion, as well as the prior Federal Circuit opinion, is complex and muddled, or at least not very clear on these three points.  I, like Justices Sotomayor and Breye r/ Ginsburg, also suggest that knowing the exact factual record relative to these three points is crucial in resolving (or at least addressing) the question of what happens when there are multiple assignments (or obligations to assign) future invention rights, such as the Stanford University Assignment Obligation and the Cetus Assignment Clause. (A fourth and crucial point not addressed by even Justice Sotomayor’s concurring opinion, as well as Justice Breyer’s dissenting opinion, is whether Holodniy had the authority or at least blessing of Stanford University to sign Cetus’ VCA containing the Cetus Assignment Clause.)

As best, all that this factual record from the Supreme Court / Federal Circuit decisions (and mainly from the Federal Circuit decision) tells us is the following:  (a) the patented technology involved methods for quantifying the Human Immunodeficiency Virus (“HIV”) by using polymerase chain reaction (“PCR”) to measure ribonucleic acid (“RNA”) from HIV in the blood plasma of infected humans who are taking drugs such as zidovudine (AZT) (hereafter referred to as the “HIV RNA assay invention”) and that Holodniy was one of the inventors listed on the patents for this HIV RNA assay invention; (b) the HIV RNA assay invention was developed in the late 1980s and early 1990s by researchers at Stanford University and Cetus; (c) Holodniy joined Stanford University in 1988; (d) Holodniy signed the agreement containing the Stanford University Assignment Obligation at a time not clearly specified but which may appears to be as early as 1988 when he joined Stanford University; (e) in February 1989, Holodniy began regular visits to Cetus over several months to learn PCR and to develop a PCR-based assay for HIV and signed Cetus’s VCA containing the Cetus Assignment Clause; (f) in December 1991, Roche purchased Cetus’ PCR business, including Cetus’ agreements with Stanford University; (g) after this purchase, Roche began manufacturing HIV detection kits employing RNA assays; (h) in May 1992, Stanford filed the patent application for the HIV RNA assay invention which resulted in the patents on which Holodniy was listed as an inventor; (i) Stanford University received government funding for its HIV research through the National Institutes of Health (“NIH”); on June 24, 1992, Stanford filed an invention disclosure for the HIV RNA assay with the NIH; and (j) on April 6, 1995, Stanford formally notified the Government that it elected to retain title under the Bayh-Dole Act to the HIV RNA assay invention described in the invention disclosure.

With this complex / muddled / unclear factual record before it, 6 Justices of the Supreme Court agreed in Stanford that Bayh-Dole doesn’t automatically cause a transfer of invention rights that might be based on federally funded research to the contractor / grantor (e.g., Stanford University); even those rights must be assigned (be it by written assignment, agreement to assign, or some “operation of law” other than Bayh-Dole) by the inventor / researcher to the contractor/grantor. But what those 6 Justices didn’t address in Stanford (despite what some in the patent blogosphere might believe) is the validity of the Federal Circuit’s ruling in FilmTec that the “do hereby assign” language in the Cetus Assignment Clause “effected a present assignment of Holodniy’s to Cetus,” and which was the basis for the Federal Circuit holding in the Stanford case that title to the  HIV RNA assay invention belonged ultimately to Roche.  And no amount of “silence” on the part of those 6 Justices as the impact of this “do hereby assign” issue can necessarily be construed as an “affirmation” that the Federal Circuit “got it right” in FilmTec, as Justices Sotomayor and Breyer both clearly pointed out. (Sotomayor concurred with those 6 Justices solely on the basis that Stanford University, in essence, “waived” that issue.)

Even in dissent, Justice Breyer (who I will admit is one of my least favorite Justices, especially when it comes to interpreting patent law issues) said he would “likely agree” with the 6 Justices that Bayh-Dole doesn’t cause an automatic transfer of invention rights to the federal contractor/grantee.  But I do give him significant credit in recognizing that Bayh-Dole does evidence a “scheme,” or as he calls it, “a hierarchy,” for how ownership of invention/patent rights is (or should be) allocated for federally funded research:  (1) first, to the funded organization; (2) next, to the federal government; and (3) last, to the employee/researcher.  In other word, Bayh-Dole provides a “pecking order” for who gets the first claim to ownership of those rights, similar to a “right of first refusal” in a license agreement.  (This was the same or similar position advanced by the U.S. Solicitor General in their amicus brief with regard to the “comprehensive framework” of Bayh-Dole “for disposition of rights in federally funded inventions.”)  In fact, Justice Breyer’s dissent goes even further in saying “I cannot so easily accept the majority’s conclusion-that the individual inventor can lawfully assign an invention (produced by public funds) to a third party, thereby taking that invention out from under the Bayh-Dole Act’s restrictions, conditions, and allocation rules.”

While this more “radical” view by Justice Breyer was obviously and expressly rejected by the 6 Justices, his further views as to the “substantive” effect of the Stanford University Assignment Obligation, versus the Cetus Assignment Clause, were not expressly rejected.  These further views were directed at the Federal Circuit holding that the Cetus Assignment Clause trumped the Stanford University Assignment Obligation because of the Federal Circuit’s FilmTec holding that the phrase “do hereby assign” “made all the difference.”  And I completely agree with Justice Breyer’s statement that “[g]iven what seem only slight linguistic differences in the contractual language, this reasoning seems to make too much of too little.”  (Or as I would phrase it, we’re “exalting form completely over substance.”) Breyer further observed that, at the time Holodniy executed his agreement to assign in 1988 (before the FilmTec case), “patent law appears to have long specified that a present assignment of future inventions (as in both [Clauses] here) conveyed equitable, but not legal, title” (citing Curtis’ A Treatise on the Law of Patents, as well as a Comment in the Yale Law Journal).  As Breyer saw it, the Cetus Assignment Clause, as well as the Stanford University Assignment Obligation, “would give rise only to equitable interests” in the HIV RNA assay invention.  And “as two claims in equity,” because the Stanford University Assignment Obligation came chronologically before the Cetus Assignment Clause, and because Stanford University had subsequently obtained an assignment of the HIV RNA assay invention from Holodniy after it existed, that “should have meant that Stanford [University], not Cetus, would receive the [invention] rights.” (It’s not exactly clear from Justice Breyer’s dissent whether he was postulating this statement based on the Stanford University Assignment Obligation being “factually” first in time, or “theoretically” first in time, relative to the Cetus Assignment Clause).

Justice Breyer also characterized the Federal Circuit holding in FilmTec as “adopting the new rule” such that “Cetus must win.”  He also noted that the Federal Circuit “provided no explanation for what seems a significant change in the law” in FilmTec, and also “did [not] give any explanation for that change” in the Stanford case.  Instead, Justice Breyer saw the Federal Circuit’s FilmTec rule as “undercut[ting] the objectives of Bayh-Dole” and “remain[ing] a technical drafting trap for the unwary” (views with which I also completely agree).

I would further add that the Federal Circuit’s “formalistic” FilmTec rule (as applied by them in the Stanford case) in allocating invention rights between competing ownership claims (one of which is by a university subject to Bayh-Dole obligations) is also inconsistent with (or at least potentially frustrates) other provisions of Bayh-Dole.  That’s not a new view by me, but one previously expressed in my earlier article cited above, in particular with respect to frustrating what is obviously intended by 35 USC § 202(c)(7).  35 USC § 202(c)(7) expressly prohibits a nonprofit organization, like Stanford University, from assigning rights in federally sponsored research to third parties (like Cetus or Roche) without approval of the respective federal agency. But the “formalistic” FilmTec rule now “opens the door” to “sham” transfers of invention rights in federally funded research by universities to entities which are otherwise prohibited by 35 USC § 202(c)(7) without prior federal agency approval.

Some will say that such a “sham” transfer may be challenged by affected third parties.  But I suggest that such “sham” transfers likely cannot be challenged by third parties, even if the federal agency later discovers the “sham” transfer, but chooses to do nothing about it.  Why do I say that?  Because the Federal Circuit’s 2007 decision in Central Admixture Pharmacy Services, Inc. v. Advanced Cardiac Solutions, P.C. basically says that, when the provisions of Bayh-Dole are violated, only the appropriate federal agency involved (in its complete discretion) may challenge such violations.  Put differently, if the appropriate federal agency chooses not to challenge such a “sham” transfer, I don’t see how even an affected third party can do so in view of the holding in Central Admixture.

Besides 35 USC § 202(c)(7), what the “formalistic” FilmTec rule (as applied by the Federal Circuit in the Stanford case) also potentially undermines is the funding federal agency’s stake in any invention rights created by that federally funded researched.  In particular, under 35 USC § 202(c)(4), if the contractor/grantor elects to take title to invention rights created by federally funded researched, the federal agency gets (at minimum) “a nonexclusive, nontransferable, irrevocable, paid-up license to practice or have practiced for or on behalf of the United States any subject invention throughout the world.”  But under application of the “formalistic” FilmTec rule according to the Federal Circuit in the Stanford case, even when the contractor/grantor does get an obligation to assign those invention rights to itself (as it must do to “elect to take title” under Bayh-Dole), the federal agency’s potential “nonexclusive license” in those rights is now squarely in jeopardy.

Admittedly, the factual situation in the Stanford case of competing claims to ownership in invention rights where one of the competing parties (like the Stanford University) is basing that claim on federally funded research may not be typical and may be somewhat unusual.  After all, Bayh-Dole has existed for over 30 years and only one other instance (noted in my article cited above) like the Stanford case has occurred.  But the mere possibility of the application of the “formalistic” FilmTec rule is causing some extreme “angst” among many universities who have existing contractual clauses (or policies) which create (similar to the Stanford University Assignment), at most, a present obligation to assign future invention rights.  And contrary to what others have said, think or believe, the question of the continuing validity of the “formalistic” FilmTec rule remains open for challenge before the Supreme Court. The Supreme Court has yet to squarely rule on whether the Federal Circuit “got it right” in FilmTec.

*© 2011 Eric W. Guttag


Warning & Disclaimer: The pages, articles and comments on do not constitute legal advice, nor do they create any attorney-client relationship. The articles published express the personal opinion and views of the author as of the time of publication and should not be attributed to the author’s employer, clients or the sponsors of Read more.

Join the Discussion

10 comments so far.

  • [Avatar for Anon]
    July 19, 2011 09:24 am

    And I’ll also agree with you that using the “do hereby assign” language in future gives you more comfort here.

    Thanks Eric, but I am not looking for “more comfort.”

    We are lawyers. We use words to impart specific meaning. To ignore the difference between an agreement to agree, as it were, and an actual present assignment – or even worse to pretend that such is not important as Breyer does, belittles the actual “art” of using apt words that is our profession.

    Yes, FilmTec does contain a convoluted fact pattern, with multiple trips up and down the court chain. But the point of language under discussion between us it not convoluted at all. I maintain that this point of difference that I am discussing with you is very straight forward. If you maintain the opinion that the language does not matter, you are doing a disservice to your clients. Your comfort with this should be on your mind – not your concern with my comfort.

  • [Avatar for EG]
    July 19, 2011 05:16 am


    I hear you and thanks for your comments. The point I was trying to make is that the Federal Circuit in the Stanford case puts quite a bit of gloss on the ruling in FilmTec. There’s quite a bit more in the factual situation in FilmTec (which is pretty complex and often hard to follow from how it’s described by Judge Plager who wrote the opinion) than you would think there was for how “matter of fact” it’s treated in Judge Linn’s opinion.

    As far as being “foolish,” let me just say that it isn’t just Breyer who see’s issues with the FilmTec ruling (or how it’s treated by the Federal Circuit in Stanford); there’s also Justice Ginsburg (who joined his dissenting opinion), as well as Justice Sotomayor (in her concurring opinion); the other 6 Justices simply didn’t address this issue. We’ll just have to see how this plays out if another Stanford situation comes along (which is rare, so the odds aren’t great). And I’ll also agree with you that using the “do hereby assign” language in future gives you more comfort here.

  • [Avatar for Anon]
    July 18, 2011 11:56 am


    I took your advice and re-read the FilmTec case (noting too the proper distinction given to a present assignment there).

    I am not sure what point you are trying to make. My reading of explicit contract terms still does nto depend on the FilmTec case to understand that a real distinction exists between an obligation or agreement to assign and an assignment in the present tense (the importance of “apt words” is known to be critical in contracts).

    To me, this is basic contract stuff, no matter how much Justice Breyer would llike to hand waive it away. By all means you are wlecome to maintain your opinion, but to do so without a bais in law just seems foolish to me.

  • [Avatar for EG]
    July 18, 2011 09:11 am


    I did pull back my horses as you suggested. Go read the FilmTec case at: . There’s more there than meets the eye, including those that wrote the Stanford decision for the Federal Circuit.

  • [Avatar for Anon]
    July 15, 2011 08:06 pm


    Actually, you are quite incorrect in making the assumption that my “view that these two terms are different rests on the premise that the holding by the Federal Circuit in FilmTec (as well as Stanford) will remain valid“.

    My view on these terms being different are because they quite lliterally are different.

    It really is as simple as that. Pull back your horses and take this a step at a time.

  • [Avatar for EG]
    July 15, 2011 10:18 am


    Thanks for bringing up the “conflict” and “inconsistencies’ in Federal Circuit decisions that treat certain state law issues (e.g., contracts and assignments) as being governed by “federal common law” when patents are involved, as well as O’Malley’s/Newman’s dissent from the denial of rehearing en banc in the Abraxis case. This goes right to my point about the Federal Circuit’s FilmTec ruling on the effect of the phrase “do hereby assign” being on very shaky ground, especially if SCOTUS ever gets ahold of a Federal Circuit decision which is based on this FilmTec ruling.

  • [Avatar for EG]
    July 15, 2011 10:03 am


    You’re entitled your opinion as I’m entitled to mine. But your view that these two terms are different rests on the premise that the holding by the Federal Circuit in FilmTec (as well as Stanford) will remain valid. That’s still in doubt if SCOTUS ever gets ahold of a case with the same/similar situation as in Stanford.

    And Breyer’s argument that the Federal Circuit holding in FlimTec essentially exalts form absurdly over substance isn’t “specious” in my opinion (which I’m again entitled to). To say that you can effect an immediate legal transfer/assign of rights (as opposed to an equitable obligation to transfer/assign as Breyer suggests) rights that don’t presently exist is in my bizarre. Again, you’re entitled to your opinion but so am I to mine.

    Of course Stanford could have drafted their agreement with Holodniy to use Cetus’ “do hereby assign” language and could certainly do so now in future agreements (that still leaves all those existing agreements of Stanford and others to consider). But in my opinion, Roche was simply “lucky” that this phrase in a “boilerplate” assignment clause in what was essentially a “boilerplate” confidentialty agreement (which was signed by someone who it isn’t clear from the factual record had authority or approval from Stanford to sign) won the day for them. Given that SCOTUS doesn’t give the Federal Circuit view much credence when it involves more than patent law, I would be reluctant to ever put all my eggs in the FilmTec basket.

  • [Avatar for Paul F. Morgan]
    Paul F. Morgan
    July 15, 2011 09:58 am

    The Jan./Feb. 2011 ABA IPL “LANDSLIDE” magazine starting on p.24 has an extensive legal research article on inconsistencies and conflicts in Fed. Cir. cases with basic Sup. Ct. principles on the [normal, unless Federally essential] utilization of state law versus “federal common law”. Board of Trustees of the Leland Stanford Jr. University v. Roche Molecular Systems, Inc., 583 F.3d 832, 841 (Fed. Cir. 2009) is cited. However, this article was apparently written before cert was taken by the Sup. Ct. in that case.

    Then we additionally had another relevant Fed. Cir. decision on this issue:
    “Judge O’Malley Dissents From Denial of En Banc Rehearing in Case Involving Patent Assignment Agreement — Abraxis Bioscience, Inc. v. Navinta LLC 2009-1539 — On March 14 the Federal Circuit denied Abraxis’s petition for a rehearing en banc by vote of 8 to 2. Judge O’Malley, joined by Judge Newman, filed a 19-page dissenting opinion, arguing that the 3-judge panel created federal common law to govern assignments of existing patents. She said the decision was in conflict with Supreme Court precedent restricting judicial preemption of state law. According to O’Malley, the panel ignored New York state law when interpreting the contract. The panel decision, issued November 9, 2010, ruled that a contract to assign existing patents was insufficient to give standing to sue. O’Malley said the panel imposed burdens on parties seeking to transfer patent rights.” [IPO NEWS 3/15/11]

    Amazingly, I found that the above Fed. Cir. Abraxis decision mentions, once, their prior Stanford case, but without noting that the Supreme Court had taken cert on it! [Footnote 3 of the dissent does note the above “LANDSIDE” article.]

  • [Avatar for Anon]
    July 15, 2011 08:31 am


    Watch your conflation of “a present assignment” with “a contractual obligation to assign.”

    You are treating the two as if they are the same.

    They are not.

    You did this at least three times in your arguments, and each time you do so, you weaken the point you are trying to make. Verb tense does matter and has, as Justice Breyer may wish otheriwse, a real and substantial meaning. Knowing (and using) such terms are why we are paid handsomely.

    As to any alleged “technical drafting trap for the unwary,” that is a specious argument given the number of times this issue has surfaced in the past thirty years.

    The basic bottom line here is that Stanford drafted and executed a poorly written agreement and must live with the consequences. This is neither a shocking nor inequitable result, and having to live with the consequences of poorly drafted contracts happens all of the time.

  • [Avatar for Michael F. Martin]
    Michael F. Martin
    July 14, 2011 01:29 pm

    “The inventor of a new and useful improvement certainly has no exclusive right to it, until he obtains a patent. This right is created by the patent, and no suit can be maintained by the inventor against any one for using it before the patent is issued. But the discoverer of a new and useful improvement is vested by law with an inchoate right to its exclusive use, which he may perfect and make absolute by proceeding in the manner which the law requires. Fitzgerald possessed this inchoate right at the time of the assignment. The discovery had been made, and the specification prepared to obtain a patent. And it appears by the language of the assignment, that it was intended to operate upon the perfect legal title which Fitzgerald then had a lawful right to obtain, as well as upon the imperfect and inchoate interest which he actually possessed. The assignment requests that the patent may issue to the assignee. And there would seem to be no sound reason for defeating the intention of the parties by restraining the assignment to the latter interest, and compelling them to execute another transfer, unless the act of Congress makes it necessary. The court think it does not. The act of 1836 declares that every patent shall be assignable in law, and that the assignment must be in writing, and recorded within the time specified. But the thing to be assigned is not the mere parchment on which the grant is written. It is the monopoly which the grant confers: the right of property which it creates. And when the party has acquired an inchoate right to it, and the power to make that right perfect and absolute at his pleasure, the assignment of his whole interest, whether executed before or after the patent issued, is equally within the provisions of the act of Congress.”

    Gayler v. Wilder, 51 US 477 – Supreme Court 1851

    According to my reading of the concurring and dissenting opinions in Roche, the underlying issue was choice of law — the potential conflict between state and federal law, in particular. I don’t think there is any real question about whether an inchoate assignment of patent rights is valid.