Red Hat global preferences

Red Hat Blog

Brookings Conference on Software and Business Method Patents Highlights Need for Reform

Last week there was a conference of significant interest to the free and open source software community on the problem of software and business method patents at the Brookings Institution in Washington, D.C. I was pleased to find that Red Hat is by no means alone in its opposition to patents on abstract ideas. There were, to be sure, some proponents of the status quo, but respected voices are acknowledging the serious problems in the current system and laying the intellectual groundwork for reform.

The conference was co-sponsored by the Computer & Communications Industry Association (of which Red Hat is a member) and Duke University School of Law (our neighbor just down the road in Durham, N.C.). A list of the two dozen scholars, lawyers, and industry representatives who participated is here.

A good number of the panelists recognized the dysfunctionality of the current patent system as it relates to software and business method patents. They acknowledged such problems for software developers as unmanageable litigation risks, the fuzzy and unpredictable boundaries of software patents, and the difficulties of identifying relevant patents. A couple of scholars noted that large software firms, among others, publicly opposed software patents in the 1980s.

Some of the harm from software patents is obvious. Do they provide any compensating benefit? There was little, if any, evidence that they encourage innovation. Although the number of software patents has exploded in recent years, one panelist expressed doubt that success in the technology area was associated with patent ownership. He observed that had the young Bill Gates been confronted at the outset with the litigation risks of tens of thousands of software patents, he might have chosen to exercise his entrepreneurial skills in a field other than software. The point, of course, is that the current system to some extent discourages innovation and entrepreneurship – a travesty of its intended purpose.

At a minimum, history suggests that patents are not a significant incentive to innovation in the software field. As I pointed out in my remarks at the conference, the Federal Circuit case law finding software to be patentable mostly dates from the mid-1990s, and the software patent explosion has occurred in the last ten years or so. However, a great deal of software now in everyday use was created earlier. Free and open source software programs such as GNU Emacs, GCC, and Linux date from the 1980s and early 1990s. Some of the most widely used proprietary software programs, like Lotus 1-2-3, Microsoft Word, and Oracle were released in the early 1980s. There’s no reason to think that the developers of those and other successful software programs would have been more innovative if they could have obtained patents.

It is theoretically possible that some software developers today are motivated by the hope of a new patent, but the likelier impact of our current patent system is to distract developers with anxieties about being sued over preexisting patents. We know beyond question that the the incentives of the patent system are not encouraging free and open source software developers. A patent entitles the holder to exclude others from making, using, and selling an invention. FOSS developers don’t want to exclude others this way – they want to share their code – and so FOSS developers in principle have no interest in obtaining patents.

So how do we address the software patent mess? Some at the Brookings conference argued for improvements to patent quality through more rigorous application of existing rules in areas such as obviousness and enablement. The simplest solution, though, would be to return to the system that we had not so long ago, in which abstract software is deemed outside the range of patentable subject matter. The court of appeals decision last October in the Bilski case points to this as a viable possibility. Court decisions in the coming months applying the Bilski test could close the door on software patents. In the meantime, thanks to the CCIA, Duke, and Brookings for promoting dialog on these important issues.