The US patent system is under fire from all sides. While open sourcers hate it just for existing and for granting philosophically distasteful tech monopolies to giant companies, big business attacks it for being too slow and backlogged, while other businesses claim that it awards patents too easily and broadly.

The giant organisation admits that it is in trouble: Commissioner for Patents John Doll told OUT-LAW Radio that if he closed the doors to new patent applications tomorrow it would be two and a half years before the current backlog is cleared.

What to do? Well the US Patents and Trademarks Office (USPTO) has taken a bold step. Kind of. It has embraced the mania of the day, the theory of the wisdom of crowds, and will attempt to harness all of our knowledge to help improve the quality of patents.

The new scheme relates to prior art, which is the name given to evidence that a patent application is not new, that someone had invented or patented something similar in the past.

The USPTO has signed up to a scheme invented at the New York Law School which throws open the search for prior art to the world. You can now look at a patent application online and submit anything you know of that would count as prior art and invalidate the patent application.

This is a fantastic idea, solving a number of problems. It reduces the workload of, and pressure on, beleaguered patent examiners by providing them with relevant, free information. It increases the likelihood that patent hoodwinkers will be caught, which in turn reduces future patent litigation, an expensive and time consuming process that has been the death knell of many a fine company.
Best of all, it does it in an open, transparent way which involves the technological community and gives them both responsibility and credit for the smooth operation of the system, while allowing the USPTO to stay firmly in charge.

There is, though, a problem. When I said that the USPTO has embraced the scheme, that was a lie. Their's is more of the kind of half-hug you give awkwardly at parties when you don't know someone that well.

They have signed up to the system, but only on a 12-month pilot, which is fair enough, and only for those who opt into the system.

There is a legal snag, you see. Third parties are barred by law from submitting prior art to the USPTO, except for in a very short window early on in the process. So participants in the trial have to waive their right not to have prior art submitted by other people.

Imagine that you are a wily inventor or purchaser of the rights to technology looking to slip in a quick patent application to put your proto-business on sound footing. You know that your application is maybe a little too broad, or a little too close to an existing one. But there are millions of patents out there and only one examiner with just 17 or 18 hours allocated to your application. What are the chances of them finding that one little patent from 20 years ago?

Are you going to sign up to the Peer-to-Patent system? No chance.

So the plan is a good one, and could radically improve the quality of an under-fire patent system. But if the full programme adopts the same mealy-mouthed approach and allows patent applicants to opt out, it will wither and die as only the least controversial applications will pass through its filtering system.

This project is designed for the controversial, the sneaky and the dishonest. Everyone's application should be forced through the system, otherwise it is a waste of the wisdom of an already tetchy crowd.