Software Patents

by Rowan Wilson on 13 April 2005 , last updated


What is a patent?

Inventors can apply to their government for the right to own the ideas in their inventions. This right is known as a patent. Without a patent, anyone else can implement their invention. With a patent, the inventor enjoys a period during which they own the idea behind their invention, and can prevent others from employing it. They can also allow others to use their ideas for a fee or at no cost.

To get a patent, an inventor must describe their invention and in doing so demonstrate that

  • it is new
  • involves an inventive step - meaning that it is not merely a progression that would be obvious to anyone who had good knowledge of the field, and
  • it can be used as part of an industrial process - meaning any practical application outside the spheres of purely intellectual or artistic endeavour.

If the invention meets all these criteria, then the inventor can receive a patent in exchange for a yearly fee. As a result of the patent being granted, the description of the invention is made public. The patent right will lapse after a certain period set by the government, and can be allowed to lapse before that if the inventor wishes to stop paying the fees. Once a patent has lapsed, anyone can employ the ideas that are embodied by the invention. The body of lapsed patents provides a valuable resource for all who wish to work on technological and industrial solutions to problems.

Strong opinions

Strong opinions dominate both sides of the argument concerning the need for a wider acceptance of software patents. Many people will tell you that in order for small innovative businesses to survive there must be a means for them to protect the ideas they put into innovative software from unscrupulous competitors. Many other people will tell you that allowing this kind of protection will make software development the exclusive preserve of the industry’s big players who can use their vast portfolios of patents to stifle competition from small, innovative businesses.

In the sphere of education, and in the public sector in general, there is a widely-held view that software patents may hinder interoperability and access to publicly-funded data, allowing proprietary application vendors to keep data standards closed and exclusive. On the other side of the argument many point out that higher education institutions could benefit greatly from the ability to capitalise on their innovations in software in the same way as they currently do in many other areas of endeavour.

So what’s the truth of the matter?

The story so far

Back in 1973 the European Patent Convention attempted to standardise the treatment of inventions across the European Community. As well as setting out the criteria for patentability mentioned above (novelty, inventiveness, industrial application) the Convention excluded certain types of invention from being patentable. Article 52 of the Convention excludes:

  • discoveries, scientific theories and mathematical methods;
  • aesthetic creations;
  • schemes, rules and methods for performing mental acts, playing games or doing business, and programs for computers;
  • presentations of information

As well as the above exclusions diagnostic, therapeutic or surgical techniques are generally accepted as being not patentable. In 1978, the European Patent Office (EPO) was set up to strengthen the cooperation of the countries of Europe in the protection of inventions.

Of course, over the ensuing years, computers became ubiquitous, and European industry became nervous that they could not protect their research investment if it was ploughed into software development. The EPO started to interpret the 1973 exclusions more flexibly in response to this pressure from industry, and started granting patents for what were essentially computer programs by allowing claims for processes that employed programmed computers.

Meanwhile in the USA

The American position on the patentability of software has always been more permissive. Software was never excluded, and it was therefore never required to demonstrate a technical effect beyond the standard operation of a software program in order to qualify for consideration for patentability. All that was required for a program to be patentable was that it produced a useful, concrete and tangible result. This requirement was originally interpreted to mean that the operation of the program must affect the physical world in some way - like for example a control program for an industrial machine. Increasingly however the interpretation of useful, concrete and tangible was stretched to include effects that were localised to the computer itself. Software that employed a better method of memory management or completed a transaction more smoothly was considered to fall within the scope of patentability. This has led to an extremely large number of software patents being granted in the US. This in turn, some would argue, has made the business of creating software which does not infringe any patents very hard indeed.

Vicom and IBM

In 1987 a company called Vicom applied for a patent with the EPO on a mathematical process embodied in a computer program that processed digitised images. Initially their claim was rejected, as the European Patent Convention disallowed their software program from being an invention on two separate grounds - it was a software program and it was integrally a mathematical method. On appeal, however, the European Patent Office allowed the claim. Their reasoning was that

  • although mathematical methods cannot be patented, technical embodiments of them can be
  • software that implements a technical embodiment of a mathematical method (or indeed any software) should be protected if it makes a technical contribution to the state of the art in the area of endeavour it inhabits.

Put simply, the EPO found that the fact that an implementation of an invention may be in software rather than in hardware ought not to disallow it from being protected by patent law. If it does something new and useful (makes a technical contribution) that alone should be the deciding factor. To quote the judgement:

Generally speaking, an invention which would be patentable in accordance with conventional patentability criteria should not be excluded from protection by the mere fact that for its implementation modern technical means in the form of a computer program are used. Decisive is what technical contribution the invention as defined in the claim when considered as a whole makes to the known art.

This decision made it possible for a carefully-phrased patent claim to result in what is effectively a software patent. The patent was granted for the computer system when programmed rather than just the program. Thus an individual producing a piece of software that would infringe the patent if it were to be run on a computer was not technically violating the patent. Finally in 1998 the EPO revised this approach (in response to a patent appeal brought by IBM) and announced that in future software itself could be patented as long as it displayed a technical effect.


By the late 90s, then, the European Patent Office had opened the way to software receiving patentable status. In the last twenty years the EPO has granted over 30,000 patents to software inventions that satisfy their criterion of producing a technical contribution. Despite this fact there has been little litigation in the European Courts in support of these patents, and it is not entirely clear why this is. It is true that policy on the patentability of software has diverged across the European Union, with national patent regimes taking greater or lesser notice of the reinterpretations of the 1973 Convention that the EPO has engaged in. This in turn has led to uncertainty about the robustness and EU-wide enforceability of patents that include software granted by the EPO and by national patent offices. Clearly it would be preferable for the situation to be clarified.

In 2002, the European Commission drafted a European Directive on the Patentability of Computer-Implemented Inventions. This was intended to standardize how national governments dealt with software patents, by forcing them to build certain features into their patent regimes. It was also intended to harmonize European states’ patent policies with the requirements of the World Trade Organisation’s TRIPS agreement, which itself was supposed to create a worldwide accord on how intellectual property rights should be enforced.

So that’s alright then..?

Unfortunately the Directive did anything but harmonize. For starters, many people felt that the reinterpretations of the 1973 EPC exclusions should never have happened, and resented the Directive’s attempt to make these reinterpretations official. Many feared that the Directive would begin a process that would lead to a US-style of patent examination, with easily obtainable software patents and little scope for contesting bad patents.

At the heart of much of the most robust opposition to software patents is the belief that computer software is not just another area of technical endeavour, like mining or hat-making or agriculture. Software, this belief argues, is a codification of mathematical processes, and as such deserves its place alongside algorithms and mathematical methods in the European Patent Convention’s exclusions. Allowing software patents to be granted, opponents believe, would be almost the same as allowing the ownership of mathematical procedures. Some critics of software patents have proposed that the domain of software should be thought of as similar to the domains of mathematics and logic - an abstraction of real-world problems, not a distinct category of them.

Besides this typological argument for the special treatment of software, there are also many practical concerns. There is a large body of common practice in the world of software development, and the US experience has shown that it can be difficult for patent-granting authorities to correctly judge novelty against this vast common background of know-how. Members of the open source community particularly tend to feel that software patents will allow a gradual enclosure of this commons, and thus a severe limiting of the ability of the open source world to produce non-infringing software.

Proponents of the Directive maintained that the insistence on a technical contribution being present in a prospective patent should filter out all undeserving or non-novel candidates. They also point out that standardising patent law across Europe is an important and urgent aim, and that practically speaking the only way this aim can be achieved is to standardise on something close to the status quo. Anything else is certain to end up invalidating patents that have already been granted, and this, it is argued, is undesirable from both an ethical and a bureaucratic standpoint.

Strong lobbying by those who oppose liberalisation of the software patent regime in Europe has resulted in the Directive being repeatedly rejected by the European Parliament. The Council of Ministers continues to promote its adoption. The history of these battles is complex and involved. To summarise, years have passed and much concentrated lobbying by both sides of the debate has resulted in deadlock and no real change to the wording of the Directive. The anti-patent groups have shown themselves to be extremely effective in mobilising opinion and tackling well-funded opposition. It remains to be seen if, realistically, they can hold the pro-patent lobby to a draw in perpetuity.

Further developments

In March 2008 Symbian software, a mobile phone operating system developer, managed to overturn a UK decision to refuse a patent for a software invention. The Intellectual Property Office had been assessing software patentability using a four step test (derived from the judgement in the Aerotel v Telco and Macrossan’s Application case) designed to establish if a piece of software had an identifiable ‘technical effect’ outside the computer upon which it runs. If such an effect was present, according to the test, the software could be considered as eligible for patent protection. Symbian’s application had ‘failed’ the test and had been rejected on that basis. In succeeding in their appeal, Symbian established that - within the UK at least - software could enjoy patent protection even if its effects were entirely limited to improving the operation of the computer it runs on. This has effectively widened the range of software patentable in the UK.

Patents and how they affect open source development is an important topic that must be considered when adopting an open source licence. Details about the interactions between patents and open source software licences can be found in the Free and open source software and your patents briefing note.

Further reading


Related information from OSS Watch:


This document is intended to highlight areas which may cause difficulty to programmers and their managers. It was not produced by lawyers, and is not legal advice. Please consult a lawyer to resolve your licensing and intellectual property issues.