A while ago, I worked through a slide in an NIH presentation about Bayh-Dole compliance. The conclusion there was that the presentation was sloppy, lacked important details, and misrepresented the Bayh-Dole standard patent rights clause. There are other accounts of Bayh-Dole compliance from NIH. Let’s look at another, the NIH version of the iEdison “Intellectual Property Reporting Guidelines.”
In this document we have a table that shows the differences between paper and iEdison reporting, along with the various requirements. Take a look at the first requirement:
Let’s work through this restatement of the (f)(2) written agreement requirement. There’s much that is not compliant. First, the written agreement at (f)(2) of the standard patent rights clause is specific to subject inventions–not “all” inventions but only those inventions owned by the contractor, the “grantee.” The NIH ignores the distinction–and just where making that distinction would matter. The NIH goes further out on a limb:
The agreement to be obtained by the grantee/contractor organization is that the employee will abide by the terms of the patent rights clause.
The (f)(2) agreement has no such requirement. Here’s what does happen: when a contractor/grantee complies with the (f)(2) written agreement requirement (no one does–ha!), it makes each employee a party to the funding agreement. The part of the funding agreement that becomes the employee’s personal responsibility is to disclose subject inventions (we will get to this) to the contractor/grantee, to sign papers to permit patent applications to be filed, and to sign papers to establish the government’s rights in subject inventions. NIST has added an additional obligation, that the employee must assign subject inventions to the contractor/grantee, which makes garble of the written agreement.
Now a subject invention is defined as a patentable invention owned by the contractor and made in performance of work under a funding agreement. So the (f)(2) agreement covers only those inventions, not all inventions. When the (f)(2) agreement is in place, each employee is made a party to the funding agreement, with their obligations pointing to the federal government (section (f) is titled “Contractor Action to Protect the Government’s Interest”), in parallel with the contractor/grantee’s obligations.
The flow down of (f)(2) means that the obligations of the employee are those only of the (f)(2) written agreement and not the standard patent rights clause that applies to the contractor/grantee. In fact, inventors are subject to a different patent rights clause, 37 CFR 401.9, which is in turn a subset of the standard patent rights clause for businesses.
We can see that the NIH is simply wrong in its statement of the scope of requirement for employees. The (f)(2) written agreement does not require employees to abide by the contractor/grantee’s patent rights clause. Rather, it makes employees parties to the funding agreement, brings them under their own patent rights clause, and precludes the contractor/grantee from having an ownership interest in the employee’s inventions as a condition of the federal award–exactly the opposite of what the easy (and stupid) reading of NIST’s addition of an assignment requirement would suggest.
When must this (f)(2) written agreement be made? According to the standard patent rights clause, it is to be made for each funding agreement, when that funding agreement is effective. That is, (f)(2) is a requirement of the standard patent rights clause. (f)(2) is not in Bayh-Dole. (f)(2) is not statutory. (f)(2) is a requirement of a federal funding agreement patent rights clause that repeats every time a funding agreement with the standard patent rights clause comes into effect. A contractor/grantee then must have employees make the written agreement at the time those employees begin to work on a project funded at least in part by the federal government.
While it might be expedient for a university administrator to sign everyone up for the (f)(2) agreement when they are hired (“at time of employment–term of employment”), that’s not what (f)(2) requires. The NIH is simply wrong, then, that (f)(2) has anything to do with “employment” by the contractor/grantee. For grants to universities, university faculty are “released” from university duties to work on the federal grant. Faculty may be “employees” for their university duties, but they are not “employees” for their federal grant work–the university does not assign them to that work, does not direct or approve that work, that work is not for the benefit of the university, the university does not determine who may participate, what may be disclosed or published–nothing about the university-faculty relationship in a federal grant has anything to do with a master-servant relationship. Faculty aren’t employees of the university when they work on a federally funded project.
The appropriate language here for the NIH is therefore not “at time of employment” but “at time of becoming eligible to receive payment under a federally funded project.” The easy thing to do is to have faculty investigators sign off on grant proposals that they make the (f)(2) agreement, and then require each employee to sign off on their employment paperwork to transfer some or all of their pay into the grant-funded work. Students and volunteers working on a federally funded project but not receiving pay from the grant would not be “employees” and would not be required to make the (f)(2) agreement. In fact, if the university made such participants ma,e the (f)(2) agreement, the university would be in breach of the standard patent rights clause. No problem–the standard patent rights clause is not enforced, but for fussy paperwork that doesn’t matter. Ironing shoelaces, disregarding the shoe.
Now consider NIH’s “discussion” of the (f)(2) written agreement requirement:
Grantee/contractor organizations must have policies in place regarding ownership of intellectual property, including conflict of interest issues.
Nonsense. It’s true that the NIH requires universities to have conflict of interest policies. But there is no authority for that requirement in Bayh-Dole or 37 CFR 401. Funny that. One might think that conflict of interest should figure in Bayh-Dole–both for inventors and for contractor/grantees. We would have a good time mapping what constitutes a conflict of interest relative to faculty-proposed work that’s supported by federal money awarded in the public interest. But not now. Later.
Major point: Nothing in Bayh-Dole stipulates that a grantee/contractor must have in place any policy on ownership of intellectual property. Bayh-Dole does not even cover all intellectual property. Bayh-Dole pertains only to patentable inventions owned by the grantee/contractor made in performance of work under a federal funding agreement. But worse, Bayh-Dole does not stipulate ownership. The Supreme Court in Stanford v Roche was adamant. Congress did not intend to change federal patent law under which inventors own their inventions. Nothing in Bayh-Dole, therefore, requires any grantee/contractor to have any policy on invention ownership. Bayh-Dole does not disturb invention ownership. Federal agencies are given no authority under Bayh-Dole to disturb invention ownership except in exceptional circumstances, and then by following a procedure no one can figure out how to use.
Thus, the NIH has it completely bassackwards. Bayh-Dole requires no policy on ownership of anything. Bayh-Dole is self-implementing as to ownership. Bayh-Dole is part of federal patent law. Federal patent law controls. The (f)(2) agreement requirement forces grantee/contractors to leave inventors alone with their inventions until they decide what to do–assign to the grantee/contractor, perhaps, or elect to retain title themselves under 37 CFR 401.9, and under 37 CFR 401.9 an inventor can simply blow patent rights with no obligation to the federal government or to the grantee/assignee. So screw the NIH here. They are clueless. And we aren’t even talking about what counts as good, beneficial public policy. We are only talking about the fact that the NIH cannot even get close to a representation of the law and patent rights clause. They have it reversed. They require the opposite of what the law and regulations require. And they don’t even know it.
In the next entry in the table, regarding “Invention Report,” the NIH recites the definition of subject invention, without any indication that it knows what the definition entails. Again, for a document giving guidance on compliance, this is exactly the point at which clarity is essential. NIH:
The term “subject invention” means any invention of a grantee or contractor organization conceived or first actually reduced to practice in the performance of work under a federal funding agreement (grant, cooperative agreement, contract).
They cannot even get the definition right. They don’t bother to point out that “invention” is also a defined term. It means an invention that is or may be patentable, or a plant variety protectable under the PVPA (and thus, with fun irony, federal patent law has a definition of invention that includes both patentable and unpatentable subject matter–talk about cognitive dissonance). The conceived or actually reduced to practice language is obsolete. It was used to establish first to invent priority before first-to-file became law.
Here, “conceived or first actually reduced to practice” language provides part of the scope for the government’s claim to have rights in an invention. It’s almost impossible language to decipher, really–I won’t get into it, other than to point out that “funding agreement” also is a defined term in Bayh-Dole, and extends to any assignment, substitution of parties, or subcontract, and that the “work” to be performed may be funded only “in part” by the federal government–so the “performance of work” is any project that the government funds at least some part of–chronology of the funding doesn’t matter, separate accounting for the federally funded portion doesn’t matter. Note as well that the NIH distinguishes grants, cooperative agreements, and contracts. Make sense of NIST’s new assignment requirement pertaining only to subject inventions made “under contract.” NIST, the great garbler.
NIH: “Date of submission establishes time frames for all future actions.” Not so. Date of “submission”–that is the disclosure of a subject invention by the grantee/contractor to the NIH establishes the time frame for notice of election to retain title. But once title has been retained, the time frame to file patent applications is established by the date of election to retain title. Here’s 37 CFR 401.14(c)(3):
The contractor will file its initial patent application on a subject invention to which it elects to retain title within one year after election of title
As far as I can tell, there are no other time frames for future actions that depend on the date of “submission.” Time frames for electing to retain title can be shortened for patenting bar dates, but that then is a time frame anchored from the bar date not from the submission date. Same for filing patent applications in foreign countries–within ten months of date of US application or six months from date of permission to file foreigns after a USPTO secrecy order is lifted. Nothing to do with date of “submission.” Not what NIH says at all.
Misguidance. Sloppy mess. Oh, gawd. Even the easy things to fact check they don’t bother with.
“Subcontractors retain rights to their subject inventions.” No, that’s not it at all. The subcontract paragraph (g) in the standard patent rights clause has no antecedent in Bayh-Dole. It was pulled from someone’s posterior cortex. But what (g)(1) says is this:
The subcontractor will retain all rights provided for the contractor in this clause, and the contractor will not, as part of the consideration for awarding the subcontract, obtain rights in the subcontractor’s subject inventions.
“retain all rights provided for the contractor” is not “retain rights to their subject inventions.” Paragraph (g) merely asserts the obvious–that a subcontractor is a party to the funding agreement and therefore has the benefit of the funding agreement and its patent rights clause independent of the grantee/contractor. Whether a subcontractor retains rights to subject inventions has to do with whether the subcontractor acquires any inventions and if so whether the subcontractor elects to retain ownership of those inventions.
Prime grantee/contractor organization cannot require ownership of subcontractor’s subject invention(s)
Pfah. Read and weep: “the contractor will not, as part of the consideration for awarding the subcontract, obtain rights in the subcontractor’s subject inventions.” A grantee and subcontractor can have a deal outside of the federal award and subcontract. All the patent rights clause says here is that ownership cannot be a requirement of the subcontract award. If there’s a deal outside of the federal award and subcontract, the patent rights clause has nothing to say about it. Heck, federal agencies, for STTR awards, require an upfront “IP agreement” that makes a hash of the subcontract requirement (and much else).
Think about it. If an inventor making the written agreement under (f)(2) to protect the government’s interest as required by the prime grantee/contractor is designated as a small business contractor by 37 CFR 401.9, then the (f)(2) agreement involves a compulsory subcontract by the prime. Then (g)(1) kicks in and the prime cannot have rights in the small business contractor’s–the inventor’s–inventions as consideration for the subcontract. There can be a deal on the matter, but it cannot involve the federal money. That’s it. Not that the NIH cares.
NIH gets the timing right for notice of election to retain title. But it doesn’t get the concept right in the header: it’s “election to retain title” not “election of title.” Bayh-Dole doesn’t give anyone a special right to take title. Only to retain title one already has by some other means and that still complies with Bayh-Dole.
And what about “take steps to commercialize the invention”? Not in Bayh-Dole. The standard in Bayh-Dole is “utilization” and “practical application,” which is use with benefits available to the public on reasonable terms. While Bayh-Dole does anticipate that some subject inventions will be “commercialized,” there is nothing in Bayh-Dole that requires commercialization as the only means to meet the utilization requirement.
Wherever one sees “commercialization” in Bayh-Dole, there’s also “public availability”–these are not a two-step–stating the obvious that once there’s commercialization then there will be publicly available product. Commercialization and public availability are two forms of utilization. An invention, for instance, might be made available as a standard. Or it may be licensed non-exclusively for any practitioner to use. A company might develop an in-house research tool based on a published research invention. There’s no “commercialization” of the research tool. It’s just publicly available and used.
Similarly, a lab medicine department at a medical center might implement an inventive, published diagnostic assay without sitting on its hands for five years while someone futzes around with creating a commercial product version of the assay. The lab medicine department’s implementation may be better, more specific, fewer false reports, and in multiple, improving versions compared with any eventual commercial product version. Bayh-Dole does not stipulate that lab medicine people must wait until some company has developed a commercial product version, with the delays and all the cruddy compromises.
That Bayh-Dole requires commercialization is fiction, faux Bayh-Dole. NIH is all-in on faux Bayh-Dole.
If a grantee/contractor obtains ownership of an invention and then decides it does not want that ownership, it notifies the federal agency. Amazing thing in Bayh-Dole is that a contractor can sit on an invention for nearly two years before deciding whether to keep owning what it owns, and then gets another year before filing a patent application–with extensions available. Another year and half before a patent application is published, so an invention can be suppressed for four and a half years. This, to speed innovation and public availability? Yes, you have been drinking heavily.
Look at the discussion. Total nonsense, pulled from NIH’s nether regions.
Effectively a waiver to the government. After further review the federal agency sponsor may elect title on behalf of the government. Title does not actually vest with the government until government elects to retain title.
If a grantee/contractor declines to retain title to a subject invention, the federal government may request title. Here’s the end of 35 USC 202(c)(2):
And provided further, That the Federal Government may receive title to any subject invention in which the contractor does not elect to retain rights or fails to elect rights within such times.
“May receive title” is not “elect title.” Look at 37 CFR 401.14(f)(1):
The contractor agrees to execute or to have executed and promptly deliver to the Federal agency all instruments necessary to . . . (ii) convey title to the Federal agency when requested under paragraph (d) above . . .
Yes, the contractor has to assign the subject invention to the federal agency when the federal agency requests assignment. There’s no “electing title” for federal agencies. They request title to be conveyed to them, and the grantee/contractor then signs the papers to convey that title. Why is this so difficult? “Title does not actually vest with the government until the government elects to retain title.” This wording makes it appears that Bayh-Dole is a magical vesting law. That title floats free of an invention, like a soul transported from a body, to land wherever bureaucrats decide. You elect title, or I elect title, and title floats down to us based on our election. Skip federal patent law that title to an invention vests with the inventor–and that’s all that the Constitution authorizes the federal government to do, reserve for inventors exclusive rights to their inventions.