But you can’t patent software!

A little knowledge can be a dangerous thing. This is clear to me from the reaction I sometimes get when I tell computer scientists that I am a patent attorney specialising in computer patents. Frequently the response is, “But you can’t patent software!”

That reaction is wrong, but quite understandable. The law really doesn’t help. http://www.legislation.gov.uk/ukpga/1977/37   “Computer programs” feature prominently in the list of things that are excluded from being inventions, grouped along with other unpatentable subject matter such as “business methods” and “rules for playing games.” As ever, though, the devil is in the detail, and as any patent professional dealing with computer inventions knows, all the listed exceptions are qualified by the words “as such,” which in practice means that all technical inventions can in fact be patented.

Limits on patentability

Ironically, the law on the patentability of something as modern as computer programs has its roots in very old law. Elizabeth I’s second most famous speech (not the one beginning, “although I have the body of a weak and feeble woman…” was prompted by arguments about patents. This “Golden Speech,” which contains the line, “Although you may have many greater and more mighty princes, you shall have none that love you better,” arose from a Parliamentary debate about the abuse of patent grants.  At the time, the country was in the middle of a recession and the cupboard was bare. Some things just don’t change, do they?

The controversy stemmed from the Crown’s decision to raise money for the wars against Spain by selling monopolies to rich courtiers, effectively getting ready cash up-front for the imposition of taxes on everyday goods. One of the monopolies complained of was on playing cards. Although Elizabeth’s speech effectively announced a royal enquiry into patent abuse, it evidently had little effect as regards the playing card monopoly. The government failed to take any action, only for the for monopoly to be  struck down in court two years later  as being contrary to the public interest. The court rejected the argument that raising the price of playing cards discouraged gambling amongst the poorer sorts of persons and ruled that the monopoly had been granted in error. This demonstrated for the first time limits on what could and could not be patented.

Whereas Elizabeth was able to buy off Parliament with a few well-chosen words, the Stuart monarchs had a harder time, and in 1623 the Statute of Monopolies limited patentable inventions to “any new method of manufacture.” This effectively restricted monopolies to novel inventions and formed the basis of British patent law up until 1977. All existing monopolies were called in and only those which fulfilled the new method of manufacture test were allowed to remain.

In its 300-odd years as the test for what was patentable, “method of manufacture” was a traditional British compromise. Anything which was patentable had to be a new method of manufacture. What was a new method of manufacture? Well, that was simply the sort of thing patents were granted for.

Computer programs and the modern law

In the late 1970s, however, that all changed when the U.K. joined the European Patent Convention. l  In contrast to the British common law compromise, European legal systems preferred codifying what was and was not patentable, and in the early 1960s, when negotiations on the European Patent Convention began, computer technology did not rank high amongst technologies of economic importance. At that time, it was unclear then that computer programs were anything more than of academic interest, mainly confined to the maths departments of universities. Certainly, the computer industry was nothing like what it is today.

It was against this background that the European governments agreed the list of excluded inventions. In the end, it copied them directly from another international patenting treaty, the Washington Agreement, which enabled applicants to obtain a search and preliminary opinion on an invention before having to file for protection in individual countries. In the Agreement, computer programs were exempted from subject matter which had to be searched or examined where national patent offices did not have the facilities to do so. Such an exception made sense in the punch-card era and relieved patent offices from having to wade through source code and try to work out what on earth the code was meant to do.

The exclusion of computer programs in the Washington Agreement related only to searching, but the exclusion took on a whole new meaning in the European Patent Convention, where it became a prohibition on what could actually be patented. In the absence of such a provision to would have been arguable that every new computer program would have been potentially patentable. On the other hand, it was recognised that a blanket ban on computer inventions would exclude certain meritorious inventions from protection, and it was for that reason that all of the patentability exceptions were limited to the excluded subject matter “as such.” These words were included expressly to enable the courts to decide the boundaries of what would be excluded from patent protection.

Technical contributions and technical effects

The question of what exactly constitutes “a computer program as such” was soon answered by the Boards of Appeal for the European Patent Office in a case relating to image processing. In Vicom’s Application, the Boards made “technical contribution” the touchstone for patentability, holding that a patentable invention must be novel, non-obvious and give rise to a technical effect.  A computer program would not be technical just because computerising something would make anything faster and less liable to error. Rather, in order to be a patentable invention a computer program would have to provide some kind of technical benefit which was arrived at in a new and non-obvious way.  If a computer program did all these things, then it would not be a “computer program as such” under European patent law and could validly be the subject of a patent.

Since that case, the courts have frequently been asked to consider the boundaries between what is and is not “technical.” Gale’s Application was one such early case in the computer field. Mr Gale was an incredibly clever man whom came up with a novel method of calculating square roots, and it is almost inevitable that everyone reading this article has taken advantage of Mr Gale’s invention, as it is implemented in every pocket calculator. The great benefit of Mr Gale’s invention was that the new algorithm that he had devised avoided the use of division steps which, in the computer field, are of course far harder to implement than addition, subtraction or multiplication. As a result, Mr Gale’s algorithm was much easier to implement than what had gone before.

Unfortunately, although Mr Gale was a very clever inventor, his intelligence did not extend to appreciating that there is a reason why professional patent attorneys exist. Mr Gale wrote his own patent application and filed it at the British Patent Office. The application set out his algorithm on two or three pages of description and included a throw-away reference to possibly storing a program on a CD-ROM. Because of the way Mr Gale’s invention was presented, the Patent Office rejected the case as being directed to a mere mathematical method. This was appealed through the courts, and ultimately Mr Gale’s application was rejected.  The Court of Appeal rejected an argument that recording an algorithm on a computer-readable medium such as a CD-ROM moved Mr Gale’s invention out of the realms of the abstract and into the realms of patentable subject matter. Tellingly, however, because of the way the patent application had been drafted, the Court concluded that Mr Gale’s invention did not solve any “technical problems” or give rise to a technical effect. It is clear from the later practical application of Mr Gale’s algorithm that this was not in fact correct. However, the advantages arising from the lack of division steps were not made out sufficiently in the application.

The fate of Gale’s Application is in stark contrast to that of the applicants in a test case I ran in 2008, Astron Clinica & Others.  This case arose from a change of practice at the British Patent Office, which until late 2007 had followed the EPO’s decision, on an IBM application in 1997, to allow patents including claims to computer-readable media where the recorded program solved a technical problem when it was run on a computer. In the modern world, such claims to computer programs are important as the computer program itself is the item of commerce which is bought and sold.  The British Patent Office abruptly reversed this policy in late 2006,  however, following a Court of Appeal decision relating to business methods.

The change threatened the prospects of many companies whose businesses depended on the commercialisation of new computer software. Because of the potential effect on my own clients, I put together a consortium of British applicants to challenge the change. The applicants were involved in a wide variety of technologies, ranging from printer drivers and software for micro-controllers through to image processing and software for analysing proteome data to identify drug targets. The lead applicant, Astron Clinica,   had developed software for skin imaging, which had potential application for the diagnosis of skin cancer. The one common thread was that all the applicants had applied for British patent protection in respect of novel software which solved technical problems, by for example improving printer output or achieving more efficient micro-controller control, and the software in question had the potential to be sold as a commercial product. All the applications had been rejected under the new British practice, and we took an appeal to the High Court to challenge the practice’s legitimacy. We won the case,  the Patent Office reverted to its earlier practice, and the clients got the protection they needed.

The Court of Appeal has since reaffirmed the importance of assessing the technical contribution of an invention. In Symbian’s Application, a patent relating to indirect addressing of dynamic link libraries, the Court reiterated the test for patentable subject matter as requiring one first to determine the scope of a claimed invention, and then to assess the “technical contribution” and compare it with the prior art. If the manner in which the technical contribution is achieved is not obvious, a patent will be granted.

In Symbian, the novel indirect addressing scheme made programming and accessing resources easier. In Vicom, a new image-processing algorithm processed pictures faster than the prior art. Both cases and the technologies in of the applicants involved in the Astron Clinica appeal exemplify what is meant by a “technical invention.” Paraphrasing the Olympic motto, anything which improves performance, which makes things go further, higher, faster (and to which can be added, makes things easier), has the potential to be a “technical effect,” which can be used as a hook for obtaining a patent. The key with all such inventions is that there is a technical contribution not just because a computer is involved, but rather because there is some kind of novel processing or novel data structure which gives rise to improved performance.

End thoughts

Intellectual property and patents in particular can be highly valuable assets for innovative computer companies. However, the erroneous perception that computer software is not patentable can deter many from securing and exploiting potentially lucrative monopolies that could help them build their businesses. As is clear from the case law, though, computer software is not excluded from patentability if the software in question meets the relevant criteria and, in particular, provides a technical solution to a technical problem.

The fate of Gale’s Application for what could have been an extremely lucrative monopoly shows that the stakes can be high. It also shows the value of seeking appropriate professional advice, the cost of which is, simply put, an investment with a view to a potentially much greater return where patented software meets with commercial success.

Nicholas Fox is a solicitor at Simmons & Simmons LLP focussing on electronics, telecoms and software patent litigation.  He is also qualified as a European and UK patent attorney and is the author of the practitioner’s work: A Guide to the EPC 2000. 



All images and illustrations used in our posts are licensed and have been legally acquired through official sources such as Adobe Stock