CAFC Muddies Patent Waters - With a Power Blender

There has been a lot written already about last week's Court of Appeals for the Federal Circuit "decision" in CLS Bank v. Alice Corporation and I'm sure there's more to come as people dig into the details of the full house of opinions.

For now the writing that most closely mirrors my own view is Gene Quinn over at IPWatchdog. Quinn echoes a number of other commentators who have thrown up their hands in a combination of desperation and frustration and said "seriously, guys, WTF?"

First a bit of background for those new to this game: CAFC is supposed to be the highest court in the land when it comes to patent matters. It's true that their decisions can be appealed to SCOTUS - and lots of commenters are hoping SCOTUS takes this one if only to restore some illusion of sanity - but mostly their decisions stand. Frankly, the Supreme Courts' rulings in recent patent cases don't give me any reason to believe they'd be any help here.

As the highest patent court, CAFC ought to be bringing clarity to the situation, setting out good guidelines that people can follow to know if their inventions are patentable and how to draw up valid patent claims. Whether you are pro- or anti-software patents, you want to know what the rules of the game are. This decision is like the CAFC is playing Calvinball with different rules depending on which judge you read.

We got no more than five judges of 10 agreeing on anything. Those who claim this is a victory for one side or the other are smoking something. The CAFC itself seems to have been desperate to come up with something to say en banc so they said it but who the heck knows what it will apply to. I'm sure we'll see endless interpretations and re-interpretations as lower courts struggle through this.

Some commentators have blamed the problem on there being only 10 judges and the fact that judicial nominees are being held up, including for CAFC. It's possible that if more judges had been available we might have emerged with a true majority opinion, but I tend to doubt it. The problem I see is rooted in the laws themselves, with which the judges continue to struggle. As I noted back in the Mayo decision discussion, judges seem to confuse 35 U.S.C. 101 and 35 U.S.C. 103. These two sections of the code try to specify what is patentable, but don't set out criteria anyone seems able to understand or follow. And computers just make it worse.

The 101 criteria is supposed to bar things that are 'abstract'. Back in the day when there was a nice distinction between "ideas" and "machines" this made sense. If someone had an idea they couldn't patent that; when they built a machine that operated based on their idea, the machine was the thing they went to patent. Then along came computers and pretty much everything became both abstract (programs, code, algorithms) and non-abstract (programs, code, machines) at more or less the same time. Trying to determine how 101 applies to computer programs, systems, machines, and operations is what leads to messes like this.

If I had my way I'd wave my magic wand and repeal 101 entirely. It feels like 102 and 103 are sufficient to give us guidance and I can't imagine that simplifying the laws would lead to worse outcomes than we have today. This might not make happy those people who want to ban software patents and think 101 is required for that, but I don't think we're making progress in that direction and certainly we're not making coherent progress in any direction.