Why patent trolls don’t need valid patents
Farhad Manjoo has an interesting profile of Cheryl Malone’s Article One Partners, a company which crowdsources the discovery of prior art for use in patent suits. These “amateur sleuths are stamping out patent trolls,” according to the title of the page; the headline is “How To Kill Patent Trolls.”
Which is why it’s surprising that there’s no indication in the article whatsoever that patent trolls are even being harmed, let alone killed, by the actions of Article One. Instead, the whole piece is based on a rather rocky syllogism. If you can find prior art, goes the argument, you can kill a patent troll. Article One finds prior art. Therefore, Article One kills patent trolls. Here’s Manjoo:
Patent trolls should be easy to defeat. Say a company tries to enforce some ridiculous claim, as in the recent case over a 1994 patent covering the entire “interactive Web.” It doesn’t seem hard to invalidate a patent that broad. All you need to do is find descriptions of that invention that date back to before the patent was filed.
The problem is that searching for old inventions is really difficult.
Does Manjoo really believe that “all you need to do” is find prior art, and the court case automagically disappears in a puff of smoke? It would be wonderful were that the case. But the real world, sadly, behaves differently.
Probably the most famous patent-troll case in recent years was the one where a troll named NTP managed to extract $612.5 million from Research in Motion. That case covered five different patents: of the five, the U.S. Patent Office had given “non-final” rejections to all of them, and had issued a final rejection to one, when the case was settled.
RIM had discovered prior art for all of the patents that NTP was suing over — but that didn’t really help them at all. The problem was that the patents had already been awarded to NTP, which meant that NTP was within its rights to sue RIM for as long as it held those patents. Once RIM found out what NTP was up to, it could and did challenge the patents at the U.S. Patent Office, which has a procedure for such things. But the U.S. Patent Office is an entirely separate entity from the U.S. District Court, where judge James Spencer made it very clear that his job was to rule only on whether RIM was violating NTP’s patents, and not on whether NTP’s patents were properly granted. Had RIM not settled the case, the court could and probably would have shut down the entire BlackBerry service.
RIM, of course, offered to post a substantially greater settlement if it could get the money back were NTP’s patents deemed invalid; NTP, naturally, rejected that offer. And challenging patents at the U.S. Patent Office takes time; if you’ve already been sued by a patent troll in U.S. District Court or just about anywhere else, it’s almost certainly too late at that point to look for prior art, take it to the USPTO, get the patent invalidated, and win the case that way. Meanwhile, it’s pretty much impossible to keep tabs on every patent awarded to a possible troll, and try to challenge those patents at the USPTO on the off chance that if you don’t, those patents might be used against you.
So while Article One is surely doing God’s work out there, I think it’s massively overoptimistic to believe that they will make so much as a dent in the patent-troll industry. What they’re doing might well be necessary to kill patent trolls. But it’s very, very far from sufficient.
Update: As commenter rootless_e points out, the jury in the NTP vs RIM case did find against RIM’s claims of prior art. Largely because RIM’s courtroom strategy was unbelievably boneheaded. But the fact remains that RIM was forced to settle the claim before the USPTO could invalidate NTP’s patents.