Recent decisions by the US courts and Patent and Trademark Office have suggested a move towards the rejection of software patents.[1] This is likely an attempt to resolve the issuance of junk patents, the difficulty in developing software that doesn’t infringe existing patents, and the resulting patent-troll litigation.[2] Irrespective of the current problems, software should remain patentable. The solution lies in a test that is better aligned with the purpose of patenting.
The purpose of a patent is to diffuse innovation to society [3]. It is based on the idea that people will innovate and disclose if they receive an exclusive right to their innovations. However, if the claim of invention is too wide and includes basic tools of scientific and technological work, then this will hamper other peoples’ efforts and discourage innovation.[4] It follows that a balance must be struck between providing enough protection to encourage initial innovation while not providing too much protection to discourage and inhibit future innovation.
It is likely that US and Canadian courts were trying to strike this balance when they held that “Phenomena of nature, mental processes and abstract intellectual concepts are not patentable” because they are basic tools of science and technology [5]. While this test may suffice for the other patentable areas, it possesses particular problems for software.
Software uses sequences of mathematical operations and laws of nature to solve real-world problems and thus sits somewhere between an abstract mental process and a physically tangible machine. Practical examples demonstrate, however, that some software should be patentable. Take for example RSA encryption. Based on mathematical manipulations, the claims were practical, concrete and had limited application in electronic signal cryptography. The patent was not for a basic tool of technology which would limit future innovation.[6]
The fact that software can also be copyrighted may lead some to believe that a patent is redundant.[7] The reason software needs copyright protection is because it, like the other copyrightable works, can easily be duplicated due to advances in technology. Copyright provides a barrier against end-users and third-parties copying and benefiting from a specific piece of software in its entirety without compensating the owner. Conversely, patents restrict a different audience who wish to copy a method of solving a problem and incorporate into their own software for re-sale. Both forms of protection solve different issues and are necessary.
The most compelling reason in favour of software patents, though, is that innovations in software are locked away in compiled code and behind web servers. Because of the inherently “hidden” nature of software, if it weren’t for patents and the public disclosure requirement [8], diffusion of these innovations into society would be limited. This may be an even stronger policy argument for software patents since in most other areas observing or disassembling the end product usually reveals the innovation anyways.
While the USPTO and courts work to cleanup the current software patent problems, it is important they remember the underlying policy reasons and the balance that must be maintained.
[1] Timothy B. Lee, Patent Office finds voice, calls for software patent sanity (July 28, 2008), online: http://arstechnica.com/news.ars/post/20080728-patent-office-becoming-a-voice-for-software-patent-sanity.html
[2] Timothy B. Lee, Analysis: appeals court unlikely to fix software patent mess (May 15, 2008) , online: http://arstechnica.com/articles/culture/clean-up-patent-mess.ars
[3] Canadian Intellectual Property Office; A Guide to Patents: Introduction, Who We are (at September 29, 2008) http://www.cipo.ic.gc.ca/epic/site/cipointernet-internetopic.nsf/en/wr01090e.html#sec3
[4] Fox on Canadian Patent Law and Practice, 4th ed., at pages 195-96
[5] Gottschalk v Benson , 409 U.S. 63 (1972), online: http://caselaw.lp.findlaw.com/scripts/printer_friendly.pl?page=us/409/63.html
[6]http://www.google.com/patents?id=daEsAAAAEBAJ&printsec=abstract&zoom=4&dq=4405829#PPA1,M1
[7] Malcolm Mooney (July 27, 2008), in response to John Duffy, The Death of Google's Patents? (July 21, 2008)), online: http://www.patentlyo.com/patent/2008/07/the-death-of-go.html#comment-124082544
[8] Patent Act, R.S.C. 1985, c. P-4, s. 27, online: http://www.canlii.org///ca/sta/p-4/sec27.html