A major law firm’s new FAQ on Bayh-Dole, 2: Subject inventions

We are working through a sparkling fresh FAQ on Bayh-Dole by a major law firm, apparently directed at small companies fool enough to consider taking SBIR funding from the federal government.

Now we must get into the deep mud. The FAQ fumbles toward an all-important question, but doesn’t get there because the FAQ does not grasp the architecture of Bayh-Dole and instead of making anything clear, the FAQ repeats phrases from the law without indicating that the words are fraught with meaning. This would be the place to demonstrate that there’s a lot going on, that a great IP attorney could guide you, fool company, through it all, and don’t worry the details because that’s what the great IP attorney is there for. But alas, no.

FAQ again: What inventions are covered by Bayh-Dole provisions?

This is a poorly framed question. Bayh-Dole’s provisions dictate to federal agencies what patent rights clauses they must use unless they justify some other patent rights clause in a given circumstance. The provisions that matter are of the patent rights clause that eventually ends up in any given funding agreement. Anyway, here’s the FAQ’s answer:

Any invention of the contractor conceived or first actually reduced to practice in the performance of work under the grant or contract is considered a “Subject Invention” under the Bayh-Dole law.

This answer is so unhelpful, merely reciting the law and even getting that wrong, burying the specially defined terms–invention, of, performance of work, and the like. Bayh-Dole introduces an idiosyncratic definition of invention–(i) is or may be patentable or otherwise protectible under 35 USC or (ii) a plant variety protectable under 7 USC 2321 et seq. Then Bayh-Dole defines subject invention–where the FAQ answer plops down–as an invention (see above) of the contractor. That is, a subject invention must be owned by a contractor, by a party to the funding agreement. Not owned by a contractor? Not a subject invention. End.

Then we get the weird expansion of scope “conceived or first actually reduced to practice” “in the performance of work under a funding agreement” (not just a grant or contract–could also be a cooperative agreement). We have to deal with these three elements. It gets ugly. Best advice for a company needing to read a FAQ such as this one here is to get a good IP attorney–preferably one that knows Bayh-Dole better than the ones here.

First point. If an invention has not been both conceived and reduced to practice, it is not patentable. Both conception (mental image of all elements) and reduction to practice (demonstration that the invention works for its intended purpose, either constructively by fully documenting the invention or actually by testing) must be achieved for an invention to become patentable.

Second point. Conception and first actual reduction to practice are obsolete terms and misapplied by Bayh-Dole anyway. These terms had to do with establishing who was first to invent an invention where there was any dispute on the matter. All this went away with first to file under the AIA. No matter, they are still in Bayh-Dole, where they are used to define scope rather than priority. So strangely incompetent.

Third point. Scope. This is what matters. And it isn’t easy stuff. And NIST and other government agencies routinely get things wrong–misstate the law, ignore the Supreme Court, and whatever other hubris they want to exhibit. So, yes, your company will have to deal with these creatures just how they say–and that’s what you are getting into with your ill-advised thoughts about getting a piddly, bureaucratically awful SBIR or STTR grant.

Fourth point. To get at scope, we have to start with the seemingly innocuous “in the performance of work under a funding agreement.” It’s not what it seems. Funding agreement and work are distinct. Think of work first.  Work is an overarching project of some form. Let’s capitalize it to distinguish it from sweaty labor “work.” How is Work defined? Bayh-Dole gives no guidance. But we know this from Bayh-Dole’s definition of funding agreement that the federal government may fund only part of the Work. Thus, performance of Work “under a funding agreement” or “under contract” does not mean, merely, “paid for with federal money.” Certainly that. But it also means “any invention that gets made in the Work, regardless of whether federal money paid for that inventive work directly.

Certainly, “federally funded invention” is way too narrow to capture the scope dictated by Bayh-Dole. Bayh-Dole’s implementing regulations at 37 CFR 401.1 work through the issues–badly, as is Bayh-Dole’s modus operandi–but clearly enough to confirm that neither separate accounting nor chronology of funding determines whether an invention is outside the scope of Bayh-Dole. You can’t invent “on the side” something within the scope of the Work while getting paid by the federal government to do some part of that Work. If you want to read and weep, look at the decision in Mine Safety Appliances, where the University of Southern California tried to invent on the side in a Navy contract, patent the invention, license it to a company, which then sued the government for using the invention.

The scope gets worse. Not only may an invention be within scope but not directly funded by the federal government–it may be made by a party to the funding agreement that has not itself received any federal funding, such as a company or organization that takes assignment of a subject invention–that assignee becomes a party to the funding agreement, and anything that that assignee then inventions, if within the scope of the Work, is “made in the performance of work under the funding agreement.” The use of federal funds is not determinative. What matters is that this new party to the funding agreement is attempting to invent “on the side,” enabled by the assignment from another party to the funding agreement (such as your company) and so deny the government access to the inventions it has bargained for. Sucks to be that assignee. So much for how Bayh-Dole promotes commercialization, eh?

Finally we can return to “first actually reduced to practice.” This means testing to demonstrate that the invention works for its intended purpose. Here’s how the FAQ handles this:

NOTE: Even if a contractor has conceived and filed a patent application on an invention prior to receiving funding, if that contractor later “first actually reduces” the invention to practice under the funding, that invention is still considered a “Subject Invention” and the government will have rights in it. So, if you use government funds to run confirmatory experiments on your invention, your invention becomes subject to Bayh-Dole.

This is bad advice. First, contractors don’t conceive patent applications. They don’t even conceive inventions. Inventors conceive inventions. Contractors–unless individuals (there’s a whole story there, sigh)–cannot conceive. They acquire. They might file a patent application.

But that’s just the garble part of the advice. Now for the really sketchy part.

if that contractor later “first actually reduces” the invention to practice under the funding, that invention is still considered a “Subject Invention” and the government will have rights in it.

There is no support for this position. It is a bugaboo that has circulated among university licensing officials for decades, like a scary campfire story about sasquatch. The issue is not whether a contractor has filed a patent application but whether that application is complete with regard to demonstrating reduction to practice–that the invention works as conceived for its intended purpose. The line of reasoning runs back to In re Eddie L. King, decided by the Commissioner of Patents. In that case, a civilian Air Force worker designed a better coupling system for air transport pallets, on his own time, own resources, outside his job description. The Air Force had tested a number of such couplers and none of them worked. He built working prototypes and had a patent application prepared, but didn’t file it. He provided prototypes to the Air Force for testing. The couplers worked. King filed his patent application without citing the Air Force tests and got his patent. A dispute arose over whether the government had a free right to use the coupler. Thus, In re Eddie L. King. The Commissioner (a deputy commissioner, actually) determined that the government had a royalty-free license to use the coupler because King had not filed his patent application before the testing by the government was conducted, and the testing was necessary, in light of previous failures of similar products, to demonstrate that the invention worked for the purpose the Air Force intended for it, and presumptively what King intended for it. Technically, then, the invention had not been “made” until the government conducted its tests. King by choice asked the government to participate in his inventing. Thus, there was government support for the invention, and the government had a right to a license. The Commissioner left open the question whether the government would have had rights had King filed his application before the government conducted its tests. Here’s the bit from the decision, after working through “conceived” and “first actually reduced to practice”:

Since a constructive reduction to practice did not take place
prior to the time Government resources were used to effect an actual reduction to practice on behalf of King, there is no need to determine in this case whether a constructive reduction to practice prior to the use of Government resources to accomplish an actual reduction to practice would provide any basis for holding that the invention was ‘made’ prior to those Government resources being used.

The decision notes that these distinctions have to do with interferences–disputes over who invented first under now obsolete patent law–and that “there are no corresponding competing interests in determinations under the Executive Order [10096].” That is, the use of “first actually reduced to practice” with government support to define the scope of federal interest in an invention runs skew with its purpose in patent interferences. The core teaching is that if an inventor, working to reduce an invention to practice, invites government participation in that activity–whether before a patent application is filed or while that application is pending to provide further test data to the PTO–then the invention comes within the scope of the “first actually reduced to practice” prong of Bayh-Dole’s standard patent rights clause–becomes a subject invention.

Thus, the “Basics of Bayh-Dole” advice here is bad, wrong. If the patent application is filed, it provides constructive reduction to practice. If that constructive reduction to practice is sufficient, then–while the In re Eddie L. King decision still leaves it open–the reasonable expectation is that any government claim on “first actual reduction to practice” is moot since merely building something based on a patent pending specification is not “first actually reducing” the invention “to practice” because the invention has already been “reduced” and cannot be “reduced” again a second way. If, however, the patent application is not satisfactory, and is rejected based on lack of “working example” to support enablement (35 USC 112(a)). For an extensive treatment of enablement and testing, see MPEP 2164. Thus, the issue is not whether a patent application has been filed, but whether the specification of the invention supports the scope of the claims. If so, then nothing further having to do with building a version of the invention, even if for the first time, has to do with “first actual reduction to practice.” First building may not be first actual reduction to practice.

If a company proposes government participation in testing an invention to demonstrate that it meets all the elements of what has been been conceived, then the invention has not yet been “made”–there has not been reduction to practice–and the invention, when it is made and meets all the other elements of the definition, will be a subject invention. If the company has no further need of testing and has its patent application filed, any subsequent work supported by the government is not “first actual reduction to practice” because reduction to practice has already been achieved. Whatever that subsequent “first actual” is, it is not “reduction to practice.”

Anyway, this is all the long way about to a simple observation. The advice for any client having to wade through this sewer mess that is Bayh-Dole is “get exceptionally good legal advice with regard to any inventive work that might become involved in the performance of work under a funding agreement so that it remains at best a background right and not a subject invention through inattention to the requirements of filing patent applications. Related–or background–inventions of the contractor might also be involved in a funding agreement, but there are other considerations involved, starting with 35 USC 202(f) involving compulsory licensing to third parties of a contractor’s non-subject inventions.

There is another simple observation to make at this point. If Bayh-Dole is so gosh darn inspired in its promotion of commercialization, why is it that companies are advised to watch out that their inventive work does not get scooped up as subject inventions. You would think companies would *want* their inventions to become subject inventions. But no, apparently not. There’s a good reason for the not. It’s because when an invention becomes a subject invention, it isn’t the same thing anymore. There may be good reasons not to patent or to hold a substantially weakened patent–open innovation, for instance, creating standards, avoiding IP gridlock–but if a company has other purposes, then Bayh-Dole is a stinking nightmare of bureaucracy, useless provisions, and uncertainties.

It would be so much easier if the government separated procurement from contract research services from support of science. Procurement would then state deliverables, including IP deliverables. If the government wants it, the government specifies it and then gets it. For contract research services, the government gets everything, inventions and data and all, no holding back and the government in return makes everything it gets available to all. For support of science, the government and everyone else gets a license for research uses, and anything else is up to the markets, with the provisio that the government may select awardees based on what they have done previously with inventions they have patented.

We can then consider this last bit of the FAQ advice:

So, if you use government funds to run confirmatory experiments on your invention, your invention becomes subject to Bayh-Dole.

It’s not a matter of using government funds. Or running confirmatory experiments. Or testing. It is a matter of whether “first actual reduction to practice” is within the scope of the Work for which the government provides any funding. If a company requests government funding (such as SBIR) to do the testing necessary to establish that an invention works for its intended purpose, then sure, it’s a subject invention. If a company runs other tests to ascertain quality of function, that’s not “reduction to practice” and doesn’t turn an invention already made into a subject invention.

This entry was posted in Bayh-Dole and tagged , , . Bookmark the permalink.