I've been talking to plenty of people (mostly lawyers) about the Bilski ruling on software and business method patents
while also having more time to reread the discussion in detail, and I'm going to backtrack on my original assessment. I should have known something was wrong when I wrote that CAFC may have gotten something right. They so rarely get it right, that I should have known better.
Some of the good news, I still stand by. The court clearly limited the scope of software and business method patents. It rejected using the standard set forth in State Street
in most cases. Some people are saying that since the court didn't completely reject State Street that this is not the victory I thought it was. On that, I disagree. As I said in my long post about the filings in the case, I thought an outright rejection of State Street that carves out a special exemption for software and business method patents is a bad idea
. Instead, I'm in favor of a much more stringent standard for anything
to be patentable. So, I don't have a problem with the court keeping State Street, but establishing a more stringent standard -- exactly what it did. I recognize that many folks who are focused on software patents really wanted a carveout exemption, and to them, this is a loss -- but I'd argue that it's better to have a more general standard than trying to carve out exceptions.
The part that I'm a little more concerned about is the loopholes that appear to have been left by CAFC in the decision. I was on a conference call with some of the lawyers who filed briefs (in favor of stronger patent protection), and they were spinning the ruling to be in their favor as much as possible -- but it became clear they were only doing so via loopholes. Specifically, they seem to think that as long as the software works on any device
it qualifies for patent protection under the new test. In other words, they seem to be saying that so long as you add the words "on a computer" to a claim, then you're all good. In fact, when one reporter on the call (Joe Mullin
) asked what sorts of patents this would impact, and after a moment of silence one of the lawyers blurted out that it invalidated Bilski's patent (the patent at the heart of this case) and that would be about it
. Other lawyers basically said that it would only eliminate poorly written patents
, which they seemed to define as those that failed to include that sort of "on a computer" language.
I don't think this is the actual intention of the ruling, and it will be interesting to see this tested -- but it's troublesome that already there's this huge loophole that many lawyers see. It means the court didn't do a very good job in actually establishing what the rules are for patents, and that's a problem. It will also be interesting to see if the "and on a computer" claims still get thrown out thanks to the earlier Supreme Court KSR ruling
which limited patent claims that simply combined two obvious things.
Still, in the short term, I stand by my assessment that this is a ruling in the right direction. It's not a full rejection of software or business model patents, but I think that's for the best in the long run. It's better to create proper overall rules, rather than trying to carve out exemptions and creating a patchwork of rules. However, I'm still worried about the loopholes, and how quickly lawyers with tons of patents seem ready to leap through those loopholes.