The U.S. Patent and Trademark Office has just started on a year-long experiment in peer patent review. The way the process now works is that companies and people submit patents, often with broadly-phrased claims, in hopes of getting a license on a new way of doing something. PTO examiners will push back on many of the claims (fewer than half of applications reviewed last year were immediately granted a patent). But there's been question in the patent community with how well the process is working. For example, look at the case of a company suing the Blackberry's manufacturer for patent infringement. RIM lost in court and had to pay hundreds of millions to the plaintiff, yet now there are some challenges to the patent, which may not survive. It would be too late for Blackberry, which wouldn't get to recoup the money.
So the PTO is working with New York Law School's Institute for Information Law and Policy to try a
peer review approach. Companies (including Microsoft and IBM) voluntarily provide software patent applications and anyone can comment and look for prior art that would show the invention isn't new or novel.
The idea isn't actually new; both the European and Japanese patent offices have used a different patent model from the US. Instead of doing all reviewing in-house, they grant applications and then providing a period of time in which others can challenge the basis of the patent. And software is a great place to try this - it's one of the biggest causes of patent application overload and one area where much of the community is already used to peer review in the form of open source software. But with backlogs that can run from a couple of years to five or six (particularly in biotech), it's worth a try. Almost anything is.
Labels: applications, patents, peer, PTO, review, USPTO