The AEA, Bayh-Dole, and Government Patent Claims

It has been pointed out that I should take a look at the Atomic Energy Act of 1954. The AEA is an interesting read. Bayh-Dole is at pains to carve out DOE/atomic-nuclear energy provisions, but expressly takes precedence over AEA (see 35 USC 210(a)(6)).

The earliest version of AEA stipulates that any invention made under a contract with the Atomic Energy Commission is “deemed to have been made or conceived by the Commission”. That’s an expansive claim that avoids the problem of assignment–much like current US copyright law’s “work made for hire” provision, where the author is the employer or commissioning party. The AEA’s approach gets amended in 1961 to “shall be vested in and be the property of” the Commission, with a procedure by which the Commission can “waive its claim”; otherwise the PTO is to issue the patent to the Commission as agent on behalf of the Government. In other words, inventors can file what they want, but the Government can change who the patent issues to. That is, there are no formalities as to assignment of personal property if the PTO simply changes who the patent issues to.

It appears further that by statute all inventions within the scope of the act are to be reported to the government regardless of whether funded or not, but if the Government takes a patent, then it appears also that it provides for a reasonable royalty from any subsequent licensee to be paid to the patent owner, which the patent owner may work out with the licensee, or have the royalty set by the Commission directly. And Section 159 provides a direct link to override anything in federally funded research generally.

The AEA strategy started as claiming that inventive acts take place as if by the Commission (hence, no assignment is needed, since the Commission is by law the inventor), and then was modified to directing the PTO to issue patents to the Commission regardless of inventorship, with provisions for patents already issued and taken by the Commission to allow for royalties payable to the inventors/patent owners if subsequently licensed by the Government. There appears to be something of the form of eminent domain, except that the value of the taking is only realized in the event of a license. A similar taking is possible for inventions made having national security significance, which may be taken by the government (and even then, potentially with a secrecy requirement so one cannot talk about the invention or the taking) (35 USC 181).

37 CFR 401.14(b) carries these gestures through with regard to DOE nuclear propulsion and weapons programs. There, the language is that the contractor “agrees to assign” patent rights. There doesn’t appear to be any special indication in the law how the contractor gets those rights in order to assign them. This is part of the “exceptional circumstances” breakout at 37 CFR 401.3(b)(4). Under 401.3(b), agencies still use the standard rights clause at 37 CFR 401.14(a), which is where we find the requirement for written agreements with research personnel to protect the government’s interest. That would certainly one way to do it.

Given these various approaches–

government *is* inventor

government directs patent office to *issue* the property right to the government

government *requires* assignment (from contractor, from inventor)

the law *requires* assignment

–the question is not whether the government can do any of them–apparently it can and has and still does–but which ones operate in the normal (without a finding of exceptional circumstances) operation of Bayh-Dole.

The options come down to–

i) the contractor has ownership on notice to the government

Personal vesting of patent rights is suspended, and when the contractor notifies the government of its election to retain title, it actually then and there holds title.

ii) the contractor is given permission by the government under the funding agreement to obtain ownership (or designate who should obtain ownership) via assignment;

The contractor relies on its written agreements with research personnel to require assignment of patent rights to itself or another qualified organization. The transactions follow the standards of US patent law, such as 35 USC 261.

iii) the government has ownership outright absent contractor action unless the government waives interest;

The law provides for ownership of patent rights by the government, as in the original AEA, but the contractor may pre-empt vesting by notice. If the contractor does not pre-empt vesting, then vesting takes place unless the agency formally waives its interest. Again, no assignment because no personal property rights for the inventors until the agency has taken action to release its ownership interest.

iv) the government obtains ownership by assignment from inventors or contractor upon conditions under which government may obtain title;

If the contractor does not elect to retain title, or fails in its obligations having elected to retain title, then the government may obtain ownership, and to do this, the government requires assignment from whomever holds title. That is, the government respects the personal property right, and exercises its interest as a conditional lien on title.

For i) and iii), these might be thought of by analogy with eminent domain takings, except that one may argue that the inventors here never hold the patent right as personal property, so nothing technically is taken. That title is placed with the contractor or government is different from a 35 USC 118 filing on behalf of the inventor.

Breaking down yet further, one gets to whether:

a) something legal happens that makes title to patent rights vest initially somewhere other than with the inventor–contractor or agency;

b) the written agreements in Bayh-Dole establish enforceable promises to direct the assignment of patent rights initially with inventors to other parties;

c) the law requires such action, and to enter into any other arrangement or make any other action is illegal and unenforceable (such as, a present assignment of a future invention extending to inventions made with federal support, as perhaps in Stanford v. Roche) except as permitted by 37 CFR 401.3.

My thought is that b) is the means by which Bayh-Dole is implemented. What I’m finding is a lot of belief in a) and c).

What I don’t get, however, is why there is any need in Bayh-Dole for written agreements to protect the government’s interest if the government intended to assert ownership outright, or could vest that ownership in contractors upon contractor notice without moving through inventors. If Bayh-Dole followed a), one might expect that the patent law would identify such changes in patent policy. There is an argument that since Bayh-Dole is placed in US patent law, if one can read Bayh-Dole as an a)-style scenario [title vests in contractor directly], then no other indication of a change in policy on ownership is necessary. Following a c)-style scenario [it is illegal to do other than assign as directed] would involve allowing title as a formality rather than expecting a promise to assign as partial consideration for the award of research funding. Again, could be, but I’m struck by the absence of apparatus that would signal such an intended effect of policy. The argument might be, Bayh-Dole is that apparatus, but then we are back to interpretations.

The government grants the patent right, regardless, so these various ways of doing it are all the government in a sense negotiating with itself over how, starting with the defaults to be managed by the Department of Commerce and the USPTO, and Bayh-Dole making a statement to preempt various agency positions, and to allow these to be reformed within the exception practice at 37 CFR 401.3.

Why does all this matter? Is it just navel sniffing?

Due process matters. After 30 years, it would be a tremendous step to isolate the process clearly by which title to patent vests and is conveyed in compliance with Bayh-Dole. Does Bayh-Dole pre-empt inventors’ personal property rights in inventions?

Written agreements matter. Inventor-based practices place a different interest in written agreements from that in contractor-first or agency-first practices. For research, this matters, as university researchers, in particular, may be expected to voluntarily propose the work leading to the research, and without some consideration in the activity, might choose not to bring their best work within a federal funding arrangement (or even within a use of university resources arrangement). What do research inventors get in the bargain? That would be the subject of the written agreements, beyond the protection of the government interest.

Pathways matter. If folks are taking shortcuts and not obtaining assignments when they must, they jeopardize their licensing positions. If, alternatively, they are spending resources obtaining assignments when these are not required, they are wasting those resources and creating false expectations as to ownership, entitlement, and freedom of action.

All this doesn’t argue that universities weren’t intended to have ownership–clearly they were put in the driver’s seat by Bayh-Dole. What it does open up is that there are a bunch of ways for this to happen, but that are not being exploited by US universities.

This entry was posted in Bayh-Dole, Technology Transfer. Bookmark the permalink.