By Kaitlin Mara
NEW HAVEN, CONNECTICUT – The efficacy of the patent system is not equivalent across all industries, and appears to be particularly ineffective in software, said a panel at the Computers, Freedom, and Privacy conference held at Yale University last week.

The presence of a one-size-fits-all patent system is not optimal, said Brian Kahin, a senior fellow at the Washington, DC-based Computer and Communications Industry Association, when innovation models across different technology sectors are so different.

The purpose of a patent system is to promote invention, innovation and disclosure, said Emily Berger, intellectual property fellow at the Electronic Frontier Foundation. “We are asking people to produce things for us that we didn’t have before,” she said. This begs the question, then, of whether or not software patents are necessary to get information in the public domain that would otherwise not be there, she added.

There are also areas of scientific advancement that do not deserve patents, said Konstantinos Karachalios of the European Patent Office. If something is a discovery, it should not be patented, as it has nothing to do with human innovation, he said. For instance, the human genome existed before it was found, and thus does not warrant intellectual property protection.

Mike Meurer, a professor at Boston University and author of a recent book entitled Patent Failure: How Judges, Bureaucrats, and Lawyers Put Innovators at Risk, said that patents today often impose much greater costs on their holders than they confer benefits, but that there are ways for the system to work better.

The key problem, he added, is that the patent system “doesn’t work like a property system,” as there is no clear, simple way to tell strangers where the boundaries of ownership lie.

Karachalios laid out three perceptions that, if widely held, would threaten the survival of the global patent system: one, that it is unfair and seems to make the rich richer and the poor poorer; two, that it creates barriers to access of public goods; and three, that it is seen to block technical progress. The global patent system is at risk of creating these perceptions, he said.

One Patent System for a Diverse Innovation Society?

Kahin compared the pharmaceutical model that, he said, involves lots of money spent by a small group of producers in order to get a single product to users, to the software model which he said involves millions of producers, widespread independent invention, resulting in complex products with thousands of patentable functions.

Also at issue, said Kahin, is that the complexity of modern technical products means an invention is covered by a patent portfolio, and not just a single patent, and that the increasingly large volume of prior art makes meaningful searches financially unfeasible for many innovators. The 18-month time lag before a patent application is published, coupled with uncertainty as to the reach of patent claims before a grant is issued, also leaves open many gaps in the system, providing opportunity for people to take advantage of information asymmetries, he said.

Meurer echoed the concern about uncertainty, saying that patent infringement costs are overwhelmingly incurred by people who do not realise that they are infringing and citing two famous cases: mobile communications device maker Research in Motion, which makes the Blackberry, infringing upon a technology made by technology development company NTP because it had not known a patent existed, and photography company Kodak infringing upon the patent rights of rival Polaroid because, Meurer said, the patent thicket was so complex that even a team of intellectual property lawyers failed to navigate around it.

Also, said Meurer, often lawsuits alleging patent infringement “come out of left field,” from people in industries that either do not overlap at all with the patent infringer’s area, or from industries that are related only tangentially.

Particularly obfuscating are nebulous claims that are unclear about what exactly they cover unlike physical property, in which the property is clearly demarcated. Not so with patent claims, where a lack of transparency in the application process and the recent rise in patents on process as opposed to products has led to a fair amount of confusion not only about how much of a technology is covered but also exactly who owns what.

The more functional a patent is – that is, the more it describes what something does as opposed to what it is – the more difficult it is for the patent system to work for that product, said Meurer. Thus, patents on specific chemicals tend to work fairly well, while patents on biotechnology – which tends to involve use of molecular substances called polymers, typically described by their function rather than their makeup, said Meurer – work less well. Patents on software work particularly poorly, with a dramatic upswing in the last few years in the likelihood that a new patent will face lawsuit in its first four years of existence, he said.

A lot of unpatented prior art, which does not make it into considerations at the patent and trademark office, said Berger, could add to this problem. She added that EFF is working with open source company Mozilla to crease a wiki-style platform of prior art in software they hope can be eventually used by patent examiners.

Efforts have failed so far in terms of patent reform, she said, citing the status of the Patent Reform Act as “dead in the water” in the US Congress for the time being, and asking what people interested in technology policy might do about this problem.

One thing that might help, said Berger, is to have the software community share with patent examiners its collected knowledge of prior art to help create a matching pattern between what already exists and what is claimed by patent applications. Another is to examine whether granted patents’ claims match existing innovation. Ninety percent of the time the USPTO patent and trademark office is willing to re-examine a granted patent claims if asked, she said, and this could help improve patent quality.

Karachalios also advocated examining patents after the fact. One of the chief problems of the patent system, he said, is that it looks only at what happens leading up to the patent application, but does not ask what happened afterwards to determine if the patent system is as effective as it could be.

Kaitlin Mara may be reached at kmara@ip-watch.ch.