Do academic researchers generally not worry about their work infringing on patents? Have there been cases where they wish they had?

Question

My Law SE question Instrument builder in a scientific lab w/ academic funding using a technique described in a patent for research, and for developing new technology covers both

1. Basic academic research that might employ some device, design, algorithm or process that is arguably covered by some patent held by someone else
2. Academic research that pursues something that might be patentable, that for a short period of time only used something like that described above.

The answer surprised me. It seems that at least in the US, the only situation where a patent holder can not sue someone for using something that is covered by a patent is if it were solely for amusement, to satisfy idle curiosity, or for strictly philosophical inquiry!

When one carries out research, it can sometimes be complicated, the realities are that several things need to be done in order to get a desired inquiry to yield meaningful information. We build gizmos, implement algorithms, procedures, etc. sometimes gleaned from colleagues or accumulated knowledge or by spontaneous ingenuity and invention, and to my personal knowledge, nobody runs every single thing through an exhaustive patent search and vets it as liability-free.

And yet, if the answers in Law SE are to be believed, there are probably zillions of infringements, many may be trivial, some might even be egregious (from a patent prosecution standpoint).

There are of course several reasons why viable patent prosecutions against academic researchers might not happen.

This is all new to me, so I'd like to ask:

Question: Do academic researchers generally not worry about their work infringing on patents? Have there been cases where they wish they had?

And if so, are there certain fields where it's extremely rare or much more common? For example, a Physics vs a Molecular Biology?

note: I'm focusing on academic research and let's further constrain this to publicly funded research through governmental agencies. Things get murky quickly if it is funded via industrial contract or the researcher is starting a company on the side.

Answer 1

In molecular biology we'd generally assume we were covered by "solely for amusement, to satisfy idle curiosity, or for strictly philosophical inquiry".

There are places however, where we do worry about intellectual property slightly, and that is when a supplier produces a product, covered by a patent, specially designed to be used by researchers in their research.

For example, a common protocol used in molecular biology is the Polymerase Chain Reaction. The patent on PCR itself expired long ago, but a key part of a the protocol is the use of a thermostable DNA polymerases. Companies continue to isolate and produce better, faster, more reliable and more accurate polymerases, and patent the use of them in PCR. These are then sold at a high price - the average lab will use thousands of dollars of polymerase in a year, and every lab uses them.

Once the DNA sequence of the polymerase has been identified, making the polymerase yourself is really not that difficult. As every lab uses such things, you could even imagine departments making it in bulk for all the labs in a department. But most labs don't because they worry that they might land themselves in trouble.

Answer 2

For most jurisdictions, patents grant the holder only the right to exclude commercial use. I try to answer using US law, which allows the patent holder to exclude any use of the patented invention.

Since most academic research is not for profit, the question is usually moot. Patent infringement can be constituted using a patented method, but there needs to be some tort. If I, by accident, re-invent a patented method, I can also argue that the patent is invalid because it is obvious and lacks novelty. This means that the patent holder is not incentivized to proceed with a lawsuit that might draw the patent holder's competitors into a fight to invalidate a patent.

In academia, I am not worried about patent violations. In industry, as I was told repeatedly, researchers are told not to worry about patents in their development of products and methods while they are researching and developing, because they might interpret an existing patent too broadly. Whether their proposal is violating a patent will be decided at a later time by vetting a potential product or procedure with patent lawyers. They are under more restrictions, as companies are worried about them using something licensed under a GNU or similar that disallows the company to claim patents.

Historically, there was an issue with product licenses. There was a famous case (I am quoting from memory so details might be wrong, but I was told this by one party) in which a U Wisconsin (Madison) group compared the performance of several databases. Oracle objected to the publication because publishing performance measurements was excluded according to the Oracle license and because their product did not look very good. This lead to a change in federal law that limited the right to exclude performance evaluation by researchers who had NSF grants.

Another related area, where we do need to worry, is non-disclosure agreements, which we enter often without the advice of a lawyer. A NDA can be very one-sided without being obviously so.

So, in short, in Computer Science and at a university, we tend to not worry about accidentally infringing on a patent, until we start thinking about creating a start-up and commercially using the product of our research.

I would be very interested to know what excluding others from "using" an invention actually means. The goal of the patent system is to prevent others from commercial exploitation, which means that there would be plenty of room for allowed use.

Let me make up an example. Assume that someone holds the patent for "database equi-joins by hashing" (It'a a thing). I think I have an even better way of doing it and to show that this is the case, I implement equi-joins by hashing and my new method and then evaluate both of them with many test-cases. Would this constitute "use"? I would argue that this is not the case, as I am not creating an artifact that relies on equi-joins by hashing, I am merely testing it. Of course, the patent holder in my ficticious example would want to argue that by using the patent I created great harm to the patent holder, because my publication makes the patent useless and might want to sue me for damages. I hope that if this were to succeed, federal law might be changed.

Answer 3

are there certain fields where it's extremely rare

I work in sociology, and as you can imagine patents just aren't a thing there...