Recently Reuters had a fascinating article all about the new patent thicket in pot that is appearing, thanks to legalization efforts in the US and around the globe.

With marijuana now fully legal in Canada and at least partially legalized in the majority of U.S. states, companies are rushing to patent new formulations of the age-old botanical. This year, the U.S. Patent and Trademark Office has issued 39 patents containing the words cannabis or marijuana in their summaries, up from 29 in 2017 and 14 in 2016.

And, of course, with patents come the inevitable lawsuits:

The first U.S. case is now winding its way through the courts. In a July lawsuit, Colorado-based United Cannabis Corp accused Pure Hemp Collective Inc of infringing its patent covering a liquid formulation with a high concentration of CBD, a non-psychoactive cannabis ingredient touted for its health benefits.

One of the key issues in this case and others, experts say, is whether the patent is overly broad or obvious in light of “prior art,” the existing level of science or technology against which an invention’s novelty can be judged.

Basically, there hasn’t been that much official prior art because pot was considered illegal for so many years, and no one was rushing to patent anything. And, of course, patent examiners are somewhat limited in what they’re set up to research regarding prior art, and they often rely on earlier patents and scientific articles as the basis for prior art searches. And, with pot, there aren’t so many of those.

Of course, this is actually quite reminiscent of the mess that came with software patents. For a long time, most people didn’t consider most software to be patentable (this is not entirely accurate, as there are software patents going back many decades, but many people considered it limited to a few special cases of software). However, in 1998, we got the State St. Bank case, in which the Court of Appeals for the Federal Circuit basically threw open the doors on patenting almost any software. And those doors remained completely wide open until the Alice v. CLS Bank decision in 2014 (which hasn’t totally cleaned up the mess of the State Street ruling, but has certainly helped dial back the insanity).

But, for nearly two decades after the State Street ruling, the US Patent Office was patenting software willy nilly — often despite much of it having tons of prior art or being completely obvious. A big part of the problem was that examiners, again, focused on mainly looking at earlier patents and scientific journals for evidence of prior art. But because so many people didn’t think that most software was patentable, there were very few patents to look at, and it’s pretty rare for anyone to write up the details of software in scientific journals (they just make the damn software).

That resulted in tons of broad software patents that covered things that had been done for decades or that were entirely obvious. And thus, we had huge patent thickets and massive patent fights that cost billions of dollars, caused innovative companies to go out of business, and generally were a massive tax on innovation, where almost all of the proceeds went into a few patent lawyers’ pockets. To this day it is a huge black mark on how the patent system works, and how it actually did significantly more to harm innovation than to help it.

I’m reminded of this mess in reading about the situation with patents around pot. While the situations are not entirely the same — the reasons for a lack of earlier patents are quite different — the overall impact is similar. The lack of earlier patents is creating an open field where things that have been done for years, or that are considered obvious, are still getting through the patent office with a stamp of approval. And it’s only going to create a pretty big mess with lawsuits. You would have hoped that the USPTO would have caught on by now, but apparently not.

Permalink | Comments | Email This Story

Techdirt.