Why you must get an IP assignment from your university employee

This posting is prompted partly by the important and recent decision of the US Supreme Court in the Stanford v Roche case, in which the court considered the provisions of the Bayh-Dole Act.  There are two, linked themes:

  1. What was decided in the above case, and some possible implications for the drafting of assignments between an academic and his or her employing university, particularly but not only in the US; and
  2. The practice of some UK universities to rely on section 39 of the UK Patents Act 1977 as the basis for saying that employee inventions belong to the university, so that a separate, written assignment is not required.

First, the US case.  The full name of the case is Board of Trustees of the Leland Stanford Junior University v Roche Molecular Systems Inc, et al.  The Supreme Court’s decision (number 09-1159) is dated June 6, 2011.  Chief Justice Roberts delivered the majority decision of the court (Justices Breyer and Ginsburg dissented).

The facts

The relevant facts of the case can be briefly summarised as follows:

  • 1985, Cetus Corporation begins to develop methods for quantifying levels of HIV in blood, using Cetus’ PCR techniques.
  • 1988, Cetus collaborates with scientists at Stanford to test the efficacy of new AIDS drugs.
  • At around that time, Dr Mark Holodniy (MH) joins Stanford as a research fellow and signs a copyright and patent agreement (CPA) with Stanford under which he “agrees to assign” to Stanford his “right, title and interest” in inventions arising from his employment.
  • MH works on improved method for quantifying HIV levels in patient samples.  He is not familiar with PCR, so his supervisor arranges for him to conduct research at Cetus.
  • MH signs Visitor’s Confidentiality Agreement (VCA) with Cetus, which provides that MH “will assign and do[es] hereby assign” to Cetus his “right, title and interest in each of the ideas, inventions and improvements” made “as a consequence of [his] access to Cetus.”
  • Working with Cetus employees, MH develops PCR-based procedure for calculating the amount of HIV in a patient’s blood.  This allows doctors to determine whether patient is benefitting from HIV therapy.
  • After 9 months, MH returns to Stanford.  Over the next few years, MH and other Stanford employees refine the technique, and execute IP assignments to Stanford.  Stanford secures 3 patents to the HIV measurement process.
  • 1991, Roche acquires the PCR-related assets of Cetus, including rights under VCA.  Roche commercialises the procedure.  The test kits are now used in hospitals and clinics worldwide.

The law

The Bayh-Dole Act, passed in 1980, revolutionised the ownership of intellectual property (IP) in US universities.    It provides that federal contractors (eg a university undertaking research work funded by the Federal Government of the US) may “elect to retain title to any subject invention”.  In the present case, some of Stanford’s research was funded by the National Institutes of Health, and Stanford formally notified NIH that it elected to retain title to the invention.

Legal argument, etc

  • In 2005, Stanford sues Roche, contending that Roche’s test kits infringed Stanford’s patents.
  • Roche asserts it is a co-owner of the procedure, based on MH’s assignment in the VCA.  Therefore Stanford has no standing to sue it for patent infringement.
  • Stanford claims MH has no rights to assign, because the research was federally funded, giving Stanford superior rights under Bayh-Dole.
  • District Court holds that VCA is effective,  but because of Bayh-Dole, MH has no rights to assign.
  • Court of Appeals disagrees.  CPA is merely a promise to assign, not an immediate assignment [note: see our previous blog posting on this general issue, at point 2], unlike the VCA which was an immediate assignment of MH’s rights in the invention.  Therefore Cetus acquired MH’s rights.  Bayh-Dole does not automatically void ab initio an inventor’s rights in federally-funded inventions.

Supreme Court’s majority judgment

  • US patent law starts from premise that inventor is first owner of a patent.
  • Inventor can assign his rights in an invention to a third party, eg his employer.  But, for this to be effective, in most cases the inventor must expressly assign to the employer. [NB UK patent law is different on these last two bullet points.]
  • Stanford has argued that Bayh-Dole reorders the normal priority of rights in an invention, ie title goes to employer (the federal contractor).
  • Congress has done this in other situations, eg inventions arising from contracts with the Atomic Energy Commission “shall be vested in, and be the property of, the Commission”.  This wording is unambiguous.
  • But Bayh-Dole doesn’t say that.  Instead it says that contractors may elect to retain title to any “invention of the contractor”.  This means inventions that the contractor actually owns, not inventions made by its employees which the employees own (eg because they have not assigned their rights to their employer).  Also, the phrase “elect to retain” does not automatically vest title in the contractor.  You cannot retain something unless you already own it.
  • Thus, only when an invention belongs to a contractor (eg a university) does Bayh-Dole come into play.
  • Judgment of the Court of Appeals is affirmed.

Minority (dissenting) judgment of Justic Breyer (with whom Justice Ginsburg agreed)

  • Can contractor’s (ie university’s) statutory rights under Bayh-Dole be terminated unilaterally by individual employee inventor who purports to assign rights to third party?  The likely answer to this is no, but the answer turns on points not fully briefed; the case should go back to Federal Court for further argument.
  • Bayh-Dole applies to any “invention of the contractor”.  University cannot make inventions except through its employees.  Therefore “invention of the contractor” must mean invention of employees. [IP Draughts likes this argument.]
  • Allowing an employee inventor to assign an invention produced by public funds is inconsistent with basic purposes of Bayh-Dole.  Also creates uncertainty, as potential purchaser of patents from a university cannot know whether university owns the IP or has it previously been assigned by its employees.
  • Majority decision interprets statute based on norms of patent law.  But there are competing norms governing rights in inventions for which the public has already paid.  Considered together with Bayh-Dole’s objectives, this should lead to a different interpretation of the statutory language.
  • “Agree to assign” and “hereby assign”: there are only “slight linguistic differences” between these phrases. [IP Draughts disagrees.]
  • When the CPA was signed in 1988, US law provided that a present assignment of a future invention could convey only equitable and not legal title to the invention [IP Draughts: ie like UK patent laws says today; but unlike UK copyright law].  However, US law changed in 1991, when Federal Circuit adopted new rule in FilmTec case [presumably, to make such an assignment effective to transfer legal title?].  Federal Court gave no explanation of this significant change in the law.  This new rule is a “technical drafting trap for the unwary” [we hate those at IP Draughts!], and it brings about an interpretation contrary to the intention of the parties to the earlier, Stanford contract (the CPA), and contrary to the reasonable expectations of the parties to the CPA as to how the court would interpret the relevant language.
  • Other US laws may assist in interpreting Bayh-Dole, eg Executive Order 10096, which governs the rights of Government employees.
  • These views are tentative as we haven’t had full argument on them.  The case should be remanded to the Federal Circuit for further argument.

Both of these judgments are compelling, and IP Draughts finds it difficult to choose between them.  The majority judgment depends in part on perceived drafting deficiencies in both the Bayh-Dole Act and the wording of Stanford’s standard employee IP assignment.  The minority judgment takes more of a purposive approach to interpreting these documents.  If we had been the authors of the 1988 CPA, there is a fair chance that we wouldn’t have used the words “hereby assigns” in an undertaking to assign future IP, particularly at a time when, as Breyer J points out, US law provided that such a statement would not be legally effective.

For future employment contracts between a US university and its employees, the overall message of the majority decision is clear: the contract should include the magic words that the employee “hereby assigns” its rights in inventions.

Rights of UK university employees

The legal position of researchers working at UK universities is very different to that of their colleagues in US universities.  Specifically:

  • There is no equivalent law to the Bayh-Dole Act, ie no overriding law stating that either universities or the UK Government must own the IP in Government-funded research.  There are contractual conditions associated with certain types of funding, which may be relevant in particular cases, but these do not alter the basic ownership rules under patent law.
  • Section 39 of the Patents Act 1977 provides a three-limbed code for determining whether the employer or employee owns an invention made by the employee.  In essence, the question is whether the invention was made in the course of the employee’s duties as an employee.
  • Unfortunately, the duties of an academic employee are not always clearly stated.  For example, university employment contracts typically do not state expressly that the employee’s duties include the making of inventions, unlike many employment contracts for scientists working in industry. Sometimes the university employee doesn’t even have an express duty to perform research, or the field of the research is unstated.  Express contractual obligations of this kind may seem inappropriate in the context of academic freedoms and the “semi-detached” relationship that the academic has with his employing institution.
  • In our view, best practice in the UK is always to obtain an assignment from the academic inventor, preferably before any patent application is filed, rather than rely on section 39 of the Patents Act 1977.

There are other differences in the UK, including the fact that the inventor’s employer may make the initial patent application in its own name, and the fact that a present assignment of a future patent application does may [revised since first posted – see comments] not automatically take effect as a legal transfer of title when the patent application comes into existence (unlike the position for copyright).

If a case similar to Stanford v Roche had been decided in the English courts, we suspect that the outcome would probably have been the same, although for different reasons.  However, in the unlikely event that the employee’s duties had been clearly stated in his contract of employment, and the research work in question fell within the scope of those duties, it would have been difficult under UK patent law for the outside company to argue that a personal assignment from the employee was effective to vest rights in that company.

There are many potential lessons in this case, including the desirability of checking documents before signing them.  The biggest general point may be that words such as “hereby assigns and agrees to assign” should be used in employee undertakings with respect to future inventions, in both the US and the UK.


Filed under Contract drafting, Intellectual Property, News

10 responses to “Why you must get an IP assignment from your university employee

  1. Pingback: Stanford University v. Roche Molecular Systems, Inc. and Contract Language Assigning Rights - Adams on Contract Drafting

  2. Reblogged this on IP Draughts and commented:

    This article is nearly 8 years old, but it is still relevant. “hereby assigns and agrees to assign” is a useful but imperfect shorthand expression when purporting to assign future IP.

  3. Pingback: Arbitration Might Just Be the Most Exciting Area of Contract Law

  4. Pingback: A Manual of Style for Contract Drafting Should be Required Reading

  5. Pingback: A Manual of Style for Contract Drafting Should be Required Reading

  6. Pingback: Arbitration Might Just Be the Most Exciting Area of Contract Law

  7. Pingback: Koncision » Stanford University v. Roche Molecular Systems, Inc. and Contract Language Assigning Rights

  8. I think that is an excellent summary of the case, but I would take issue with one point: commenting on the differences between US and UK laws you say “present assignment of a future patent application does not automatically take effect as a legal transfer of title when the patent application comes into existence (unlike the position for copyright)”. I always thought this was the case, but Arnold J in KCI v Smith & Nephew [2010] EWHC 1487 decided otherwise, relying on the provisions of Section 7(2)(b) of the Patents Act 1977 (see paras 60 -68). So UK law is the same as current US law on this point and it is just as important in the UK to use the “hereby assigns” wording as it is in the US.

    • Thanks, Chris. I wondered whether anyone would pick up on this good point. I agree that Arnold J did say this, and there has been some discussion of the point on the LinkedIn CIPA group (topic heading: Defective Assignments). As I commented there:
      Arnold’s judgment is interesting but not, in my view conclusive (of the point that a prospective assignment transfers legal title), for the following reasons:

      1. His acknowledgement that neither counsel cited any authority – this makes me nervous and I want to be sure that Arnold’s judgment on the point is part of the ratio decidendi.
      2. Edwards v Cook is principally concerned with who may claim priority; the validity of an assignment is merely a step on the way to this issue. Arnold goes on to discuss European practice (paras 69-72). In doing so he cites an opinion from NIcholas Pumfrey that “the assignment of the invention (which post-dated the making of the invention) was effective in law even though the assignment of the patent application was not”. This makes me more nervous. I think it is possible to construct an argument that section 7(2) PA 77 allows first ownership to vest in someone other than the inventor, but that a prospective assignment of a patent or patent application cannot vest legal title per se. To my mind section 7(2) is a somewhat backdoor route to saying, in effect, the same thing as section 91, CDPA 88 in relation to prospective assignment of copyright.
      3. Arnold shores up his defences by saying (para 69) that the prospective assignment is good to transfer beneficial but not legal title and this is sufficient for present purposes. This seems to acknowledge that his point about transfer of legal title may not be free from doubt.

      To my mind, there is enough uncertainty here to allow another High Court judge to ignore Arnold’s comments about transfer of legal title, rather than be bound by precedent.

  9. Michael Berkson

    We need to be alert to wider aspects as well.

    Many years ago, my then employers funded development work at an English university. By virtue of Patents Act 1977 s. 39, the resultant invention was jointly owned. We negotiated an agreement with the Head of Department whereby we owned all rights and would pay the University a royalty if and when the invention was commercially exploited.

    Although the University honoured the agreement, it subsequently transpired that Professor X did not have authority to assign University IPR and we should have involved the Registrar.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.