Here are two court decisions at cross-purposes. In 1997, Shaw v The Regents of the University of California held that a patent agreement should be interpreted using common law of contracts and rejected the Regents contention that its patent policy was an administrative matter subject to its discretion to change and enforce at will.
We find no merit in the University’s suggestion that, as a public employee who is employed pursuant to statute, not contract, Shaw has no vested contractual right in his terms of employment, such terms being subject to change by the University.
When a public employer chooses instead to enter into a written contract with its employee (assuming the contract is not contrary to public policy), it cannot later deny the employee the means to enforce that agreement.
We also reject the University’s argument that the Patent Policy is a mere personnel policy which it may modify unilaterally. Although the University is entitled to revise its Patent Policy, it cannot do so with respect to Shaw because of its written agreement with him.
In 2003, in Fenn v. Yale, the court held that a patent policy could be changed at will by the university, and provided that an employee did not object to the changes, the changed policy would then apply (brackets in original):
This Court has held that “employment agreements [pursuant to which an employee agrees to assign to her employer inventions she developed during the course of employment] are valid and enforceable and  do not violate public policy.”
Here, Dr. Fenn, like Chou [in Illinois], was bound by Yale’s administrative policies, including its patent policy. Indeed, as noted above, Dr. Fenn concedes that he was contractually bound by the 1975 patent policy. Dr. Fenn argues, however, that he did not assent to any subsequent patent policies, and thus, that he was not contractually bound by them. . . . Dr. Fenn has not shown that Yale’s amended patent policies in 1984, 1988, and 1989 “substantially interfered” with his legitimate expectations about the terms of employment. . . . Moreover, his continuation of work and compliance with the patent policy for two prior inventions are evidence that he agreed to the changes.
The Shaw court argues that a patent agreement forms a contract that prevents the university from later changing that contract unilaterally. And if the contract were changed unilaterally, then (in other situations, at least), courts have looked to consideration to see whether a new contract had been made. Continued employment, perhaps, might be a form of consideration for at will employees. But tenured faculty are not at will employees.
The Fenn court argues that a patent policy (no express patent agreement) is binding, and that if a university changes its policy, then its faculty have the choice to either quit or accept (even if implicitly) the changed policy. The court also assumes that continued disclosure of inventions implies acceptance of a later policy, even if both the previous policy and changed policy both require disclosure of inventions.
It’s an interesting logic. We assert something. We hire you. You rely on that assertion. Then we change what we assert. In other contexts, it is called “bait and switch.” Here, however, it appears to be “the university can change its policy unilaterally.” In other contexts, it is, “We have altered the deal. Pray we don’t alter it further” or “We are going to make you an offer you can’t refuse.”
Shaw court: you can change what you assert, but you can’t change unilaterally the deal you made with your employee.
Fenn court: you can change what you assert, and you can unilaterally change the deal with your employee.
You see the difference in the logic of the two decisions.
There are other factors at play. The Shaw case involved a public university. Fenn worked at Yale, a private school. The Shaw case did not involve Bayh-Dole. The Fenn case relied to a significant extent on a (now, post Stanford v Roche) deeply wrong interpretation of Bayh-Dole. In Shaw, the University of California required a patent agreement. In Fenn, Yale relied on the assertion that the university could claim inventions as a matter of policy, conditioned on employment. The Fenn court agreed with Yale that Fenn owed the university a fiduciary duty to report inventions and disclose their value. The Shaw court did not consider fiduciary duties and instead focused on the patent agreement that the University of California required.
It may be that a separate patent agreement displaces any implied fiduciary duties with the duties specified by contract, plus an expectation of good faith.
It seems strange that a university policy may be changed unilaterally by a university and still be considered to form a binding obligation. It’s one thing for an organization’s policy to bind its officers and employees with regard to the actions they take on behalf of the organization. It’s another thing to have a policy that requires employees to assign their personal property (an action they must do as non-employees) to the organization. It would appear that an employer must have an agreement on that matter, not merely a policy.
Here’s the Supreme Court in Stanford v Roche:
No one would claim that an autoworker who builds a car while working in a factory owns that car. But, as noted, patent law has always been different: We have rejected the idea that mere
employment is sufficient to vest title to an employee’s invention in the employer. . . .
That is, employment alone doesn’t create an obligation by an employee to assign any invention to an employer. Inventors own their inventions. All the more so for faculty at a university with academic freedom and tenure. Faculty choose what to study and aren’t assigned to research. For their research, faculty are not employees unless they expressly agree to be employees–to work for a master, to take direction from the master, to do their work for the benefit of the master.
Contractors generally institute policies to obtain assignments from their employees. . . .
That is, the policy authorizes an organization’s officers to obtain assignments from employees. To be assured of assignment, it is not sufficient to threaten to fire someone who does not assign. One has to have an agreement with regard to inventions–a contract regarding assignment. Now, if one has an agreement with regard to inventions as a condition of employment, it would appear that the scope of that agreement could be no more than the scope of what is invented in employment. Otherwise, the agreement would read: “we will fire you if you don’t assign to us inventions you make outside the scope of your employment with us–you know, stuff that’s not related to our business.”
As just noted, universities typically enter into agreements with their employees requiring the assignment to the university of rights in inventions. With an effective assignment, those inventions—if federally funded—become “subject inventions” under the Act, and the statute as a practical matter works pretty much the way Stanford says it should. The only significant difference is that it does so without violence to the basic principle of patent law that inventors own their inventions.
The Supreme Court lays out the schema that regulates ownership of inventions. A university policy might state what a university wants from its employees, but that policy, as a matter of employment, does not create any obligation for an employee-inventor to assign inventions to the university, or a claim of ownership by the university, or even–in the case of faculty–an obligation to invent, or to do all inventive work for the university or within the scope of its claims.
To secure an obligation to assign, there must be an agreement to assign–not merely an agreement to follow policy that claims assignment. If that were all that there was to it, then every organization would simply put in a policy statement (i) all employees must follow policy, and (ii) we can change policy any time we want, and (iii) oh, you have to assign anything you invent to the organization. Of course, (iii) can be added any time, once there’s (i) and (ii). This condition would then result in just what the Supreme Court said does not happen–employment is not a sufficient condition to create a claim of ownership in inventions. Thus, so also an assertion in policy that employment is a sufficient condition for a claim of ownership also is just another form of the claim that employment is sufficient, so long as there is a policy statement about it.
Further, the agreement to assign is bounded by the scope of employment. The employer has no right to claim in an agreement specific to employment any invention that is outside the scope of employment. For university faculty, employment is ambiguous. To the extent that a university can assert control over faculty research and publication, then faculty might be considered to be employed. Given statements of academic freedom and research freedom and publication freedom also in university policies, it is clear that faculty are not employees for the purposes of their research and other scholarship, even if faculty are also employees for other matters (including research that they expressly agree to do for the university, under university control). Faculty are thus at once employees and independent contractors (for university commissioned work) and independent practitioners benefiting from the dedication of university resources to their scholarship.
To summarize: employment is not sufficient to establish a claim by an employer to own inventions made by an employee. For a university to secure a claim, it needs an agreement with each inventive employee. That agreement, if it is a condition of employment, is necessarily restricted to the scope of that employment. For faculty, the scope of employment is narrow compared to the scope of potential professional activity. A policy statement might authorize university administrators to enter into an agreement with faculty. But it is the agreement, not the policy to require an agreement or to state the university’s desire to own inventions, that establishes the university’s claim.
University administrators, in an effort to avoid making such an agreement, often claim the employment agreement is a broader agreement–covering employment and any other professional activity, as if these are one and the same thing, as if academic freedom does not exist–that the university’s policy forms the conditions of employment. Perhaps, but the conditions of employment are not sufficient to establish a university claim of ownership to inventions. At least the Supreme Court ruled that way, and the Shaw court ruled that if there is an agreement, then that agreement controls, and where it incorporates policy, that policy when changed cannot change the agreement. The Fenn court says that if there is no patent agreement, then policy can substitute for that agreement, and when the university changes its policy, then if a faculty inventor does not quit, then not quitting is sufficient evidence that the faculty inventor has accepted the changed policy, or if the faculty inventor submits invention disclosures, that act must be under the changed policy and not any previous policy and certainly not voluntary.
University administrators wanting to exploit faculty IP to make a profit (or whatever public interest excuse they announce to cover their efforts) prefer to read the Fenn case and ignore the Shaw case and the Supreme Court in Stanford v Roche. Attorneys representing university administrators for some reason want university administrators to win when a matter ends up in court, and so make arguments to persuade the court regardless of whether those arguments are supported by university policy read carefully and thoroughly–and worse, regardless of whether those arguments would benefit innovation, the public, university scholarship, or even the money-making efforts of a university patent licensing program. These things are non-issues, once something gets to court. The goal is to win, by whatever means the court permits.
University administrators frequently (these days) assert that faculty owe the university everything, but administrators don’t tell the faculty this. I’ve heard this argument in the wild by ordinary senior university administrators. That is, the intention behind the policy is unstated by design, but the various elements of a policy apparatus can be brought into play whenever a university administrator wants an invention and a faculty member declines to assign it–a policy that claims all inventions (IP policy), a policy that insists that any use of resources resulting in an invention creates an ethics violation unless the invention is assigned (conflict of interest policy), that all inventions arising in research must be assigned (sponsored research policy), that any invention “relating to” one’s “field of expertise”must be assigned (non-compete policy), that any invention supported with federal funds must be assigned (the misrepresentation of Bayh-Dole policy).
It’s clear that university administrators want access to faculty inventions any way they can get them, even though doing so is bad for innovation, bad for scholarship, bad for the position of the university in society, bad for other programs of the university, bad for money-making, bad for industry, bad for public benefit–and bad for university technology transfer.