Here are some strategies for getting IP back that don’t work or avoid IP ownership.
(1) Don’t disclose an invention to the university, and file for a patent on your own. Your application will be published in a year or so. Someone will see it and ask questions. If you spin up a company someone will notice. If you license to a company someone will notice. At some point you will be asked to assign and you can fight it then, in court and with the university ready to smear your reputation and ruin your career, or you will cave.
You may be able to work the system if you have a consulting deal with a company. That depends on how well drafted and guarded the university’s policy is on consulting. Many university policies on consulting these days try to stipulate that any consulting agreement’s terms on IP are subordinate to a university’s claims. That stipulation may not hold up, but the university has more money for the legal fight than you can imagine. That’s even more the case if you also are a co-owner of the company you have set up to consult for. Then there’s likely also a policy on “deeper involvement than consulting.”
Even though you might think that the university has no claim or it would be unfair if it did make a claim on inventions that you make outside the university, with a company, say, or at least not using university time and resources, many university IP policies make a claim to any invention you make in your “field” or in the area of expertise for which you are “hired” or “employed” (ignoring that faculty are not “employed” for IP purposes unless they expressly agree to be, but for another time).
(2) Don’t, having disclosed an invention to the university, downplay or under-report its apparent value or importance with the hope that the university won’t see the value in the invention and decline to keep ownership. It doesn’t work. See the Fenn case at Yale. Nobel prize winner, dysfunctional licensing office, induced to give the technology back having bungled trying to license it, only to find out that Fenn can and has licensed it, making them look like schmucks for not doing their own diligence and angry as hornets about it, suing Fenn into the dirt, relying on a crap argument about Bayh-Dole, later shown to be crap by the Supreme Court in Stanford v Roche. But no matter. Yale won. Fenn lost. And now it’s a precedent. So you might stay silent about what you think. You might decline to answer questions–or not get around to it. But don’t lie. Don’t undersell or misdirect. It will come back to haunt.
Oh, and it doesn’t matter. Most universities don’t give back technology they can’t handle or think is worthless. Here’s the, uhum, logic. If you want the invention back, then it must be worth something, and you aren’t telling. Therefore, the only way you get it back is by taking a license that repays the university its patenting costs and pays a share of whatever you make. So all the hassle of setting up a company without the liability screen and limit on what a university routinely asks of startup companies. If you don’t appear to want the invention back, then what’s the point of wasting paperwork on it. Let it go.
Worse, if it is an invention made in work receiving federal funding (or, in the university formulation, which is wrong (sigh) “made with federal funds”), then the university will insist that it has to offer rights “back” (also wrong) to the federal government, and it is up to you the inventor to petition the federal agency to get rights at all. Maybe you will get rights, but again there has to be a special mix of the university recognizing that you see value in the invention but they think you are stark raving mad and don’t want to spend the money for a patent application beyond the provisional, and for some reason don’t want to bother with making you pay their patenting costs and owe them a royalty. Otherwise, they will make you sign an agreement in which they get everything they would have gotten if they had done the work and spent the money, except now they don’t have to–you do. Really, it takes a long time, ties you up in bureaucratic knots, and in the end makes you bitter, and so doesn’t work.
Here is one strategy that sometimes works, though it’s more about not having the IP in the first place.
(3) Develop your research project as “open.” Import open source code–especially a GPL3 or Apache variant–if you think that any inventive work will be surrounded by code. That will force your developments, at least in code, to go open upon distribution. Or, build “open” into your grant applications as a deliverable. That takes the edge off of any patent play for exclusivity or even for a demand for royalties. Takes something out of any hope for royalty sharing from the university, but it also daunts the university for taking any ownership of inventions encumbered with a commitment to open access, so maybe they don’t demand ownership and if there’s federal funding maybe they can be persuaded that under Bayh-Dole if they don’t take ownership and don’t make you a party to the funding agreement, then they have no obligation to take ownership, no obligation to report the invention (it is not a subject invention unless they own or you are a party to their funding agreement), and therefore they can leave you alone figuring that you won’t waste $10,000 filing a patent on something you will release for free.
Well, free can be very good–free means that you may have people asking you for assistance in implementing or adapting your invention to their purposes–and paying you for your time. Free means maybe a seat at the table when folks are talking about developing a standard. Free means that you may get widespread rapid adoption and lots of people citing one of your papers even as they patent improvements and applications left and right and hope to get as much money as they can screwing everyone else via their university licensing office with patent monopoly suppression of competitive practice and price-gouging if possible. But no mind–for you the citations to your paper mean everything–they show not only industry uptake but a bunch of academics following along hoping for some crumbs, if not gravy.
If you work it well, however, you can create a user community based on your open release, and push the low hanging improvements and applications out of patentability just by you and the community publishing lots of variations and suggestions for obvious extensions. And if you do come up with a great, non-obvious application, if you are smart, you will do it outside the university, without its resources, and perhaps then it will stay outside the university, patented or not. The trick is to have a co-inventor who is not bound by the university. They do the filing. You are just along for the ride as a co-inventor. What is the university going to do? Make you assign your joint interest so they can do exactly what? They can’t license exclusively, and the invention is in an open project, and so are they going to pay half the patenting costs to get nothing back? Well, that’s a hairball for them they usually avoid. But if they think for a moment you are gaming them, they will try to make you pay somehow, if only as a warning to others.
User communities are themselves intangible assets of substantial value. It’s not a value that gets turned into money directly. No, you can’t sell a user community. But you might gate access to it, sure. But better, the user community itself demonstrates a social proof that validates your technology. Similarly, creating a technology platform, or tools library, or ad hoc standard also has great value. Lots of shared expertise, common discussion of methods and directions. To be part of something like this is huge. Intangible assets everywhere. Connections. Weak ties being strong. Names for things. Position papers. Specifications. Lots of press. Chatter. A scene people want to get in on. If you have a commons of this sort, then there are plenty of opportunities to make more discoveries, to develop applications, to start companies, to make your university research have impact indirectly. Instead of trying to sell exclusivity to speculators or borgs or speculators that specialize in selling to borgs, you provide non-exclusivity and attract technologists and professional users and users who want/like to be at a technology frontier. Good stuff, this sort of non-IP intangible (NIPIA) assets.
While universities have IP policies, they typically don’t and can’t have NIPIA policies. So work the NIPIA side, and get popular, central to emerging networks, and develop those weak ties with strength. There is as much or more value in name recognition, goodwill, network externalities, and being part of a going concern, even if “only” a commons or platform or tool library. You will have opportunities to consult, skip on leave to a company for a year or more, get some equity in a startup where someone else is doing the heavy financial and legal lifting, and life can be good, and much better than sitting on your hands in a corner waiting for a university patent licensing office to make you rich. Which, as a general rule, doesn’t happen.