Back in the 1960s computer programs did not have the status of standalone products. They were written specifically for a given mainframe computer and were bundled together with the hardware as an offering to the end customers. Accordingly, computer programs had a secondary role as mere accessories to hardware. This started to change at the end of the 1960s when IBM, being under the pressure of an antitrust lawsuit, decided to unbundle the sale of software from hardware. As a result, software was priced separately and become a standalone offering. Gradually, commercial enterprises started offering competing products to the end users and this gave birth to the software market which at this point was in its infancy phase. Initially, software was a lot more popular in the field of academia as opposed to commercial enterprises and individual users. What really brought software the status that it has today and actually triggered the start of the software revolution was the appearance of the personal computer in the 1980s. After that a snowball effect followed with the release of Windows 1.0 in 1985, the first version of Microsoft Office in 1990 and the release of Mosaic – the first mainstream graphical web browser in 1993. As the market for computer programs was quickly expanding and software became widespread and commercially valuable, the question arose for the most suitable form of legal protection for computer programs. There were extensive debates and discussions throughout the 1970s and early 1980s about the different options for IP protection – copyright, patent, trade secrets, or a “sui generis” regime.
It has to be stated that software as protected subject matter has certain unique characteristics that make it difficult to find one regime that would be a perfect fit. The competition for taking the throne in governing IP protection of software during the 1970s was mainly between the patent and copyright systems. Copyright was deemed a good fit as computer programs in their source code form resemble a literary work crafted with a programming language. Patents were also seen as a good fit for the IP protection of computer programs as patents are granted for technological innovations and this was how software was seen at the time. Accordingly, many patent applications for computer programs were filed in the 1960s and 1970s. There were also proponents of a third position – “sui generis” protection, who claimed that a specific new regime for IP protection of computer programs must be created. The “sui generis” approach was premised on the underlying assumption that conventional forms of IP protection cannot be properly applied to computer software because it was unlike anything that humans have ever created. Therefore, the suggestion was to come up with a new customized statutory framework that was specifically tailored for the new technology. As an attempt for such new regime, WIPO adopted in 1977 its Model Provisions on the Protection of Computer Software, which however were never broadly adopted in the form of national laws. Finally, the debate was pretty much put to an end in 1980 when the United States adopted an amendment to its Copyright Act stipulating that computer programs are protected under copyright. Additionally, the copyright regime made it a lot easier to quickly establish an international framework for protection of computer programs based on the Berne Convention and the Universal Copyright Convention as countries only needed to stipulate in their national laws that they will treat computer programs as literary works. Gradually, consensus was reached across the globe that computer programs shall be protected under copyright while software-related inventions can be protected by patents. Therefore, today we have two main regimes for protecting IP rights in computer programs – copyright and patent.
The protection offered by the patent system is quite different in scope and intensity as opposed to the copyright regime. Naturally, patents can be a good fit for computer programs to the extent they are technical items. Pursuant to Article 27.1 of the TRIPS Agreement, patents shall be available for any inventions, whether products or processes, in all fields of technology, provided that they are new, involve an inventive step and are capable of industrial application. Patents can be granted for inventions in the form of a product and also a process. When the subject matter of a patent is a product, the patent owner has the right to prevent third parties not having the owner’s consent from the acts of making, using, offering for sale, selling, or importing for these purposes that product. When the subject matter of the patent is a process, the patent owner has the right to prevent third parties not having the owner’s consent from the act of using the process, and from the acts of using, offering for sale, selling, or importing for these purposes the product obtained directly by that process. The term of a European patent is 20 years from the date of filing of the application and generally the term is the same in most other jurisdictions. Patents are limited on a territorial basis – the exclusive rights apply only for the country in which the patent is granted. In addition to being a legal document granting an exclusive right over an invention, the patent can be viewed metaphorically as a deal between the general public and the inventor. On one hand, the inventor receives protection in the form of monopoly over his invention for a limited period and on the other, he has to publicly disclose the invention so that it can be freely used by anyone after the expiry of the protection term.
When it comes to assessing the suitability of patent protection for software, we first need to take a deeper look at the complex nature of computer programs. A computer program at its core is a set of instructions that aim at achieving a particular result or function. As these instructions are expressed in source code (lines of text in a programming language) the natural analogy to a literary work is inevitable and hints toward applying copyright protection for computer programs. At the same time, a computer program is a lot more than a mere literary work. To give a simple example, a literary work – let’s say a novel, has no functional nature, it exists entirely in the realm of ideas and cannot produce technical changes in the material world. In that sense, we can clearly see the difference with a computer program. Consequently, both components of the definition of a computer program – a set of literary instructions and a resulting function are equally important and deserving of protection. And the second component – the function, is entirely separate and distinct in its nature from the lines of code that are protected under copyright. At the same time, this second component remains unprotected under the copyright regime.
As we know, copyright law does not protect underlying ideas and principles but only their expression. A software developer might consider this to be a disadvantage, because his copyright will not preclude a competitor from creating a new work embodying the exact same idea or achieving the exact same functionality, as long as this competitor does not incorporate the copyrighted expression (source code) from the initial program. This second aspect – the functional nature of a computer program, can be effectively protected through the use of patents. Moreover, the mentioned idea-expression dichotomy is irrelevant in the realm of patent law. Therefore, patent protection is especially valuable in view of protecting the technical functionality of a computer program.
Within the EU legal landscape patent protection for software is a lot more limited option than copyright protection. First, it is the EU Computer Programs Directive 2009/24/EC that reflects the preference to copyright and second, the European Patent Convention excludes computer programs from the subject matter that can be protected through patents. According to Article 52(2)c of the European Patent Convention, a computer program cannot be a patentable invention. The reasoning behind this provision is that computer programs are seen as mere instructions (lines of text) and by default they are deemed to have no technical character. This lack of technical character automatically means that they cannot be an invention which is the first basic requirement for obtaining a patent. However, this exclusion is not absolute and what is key here are the words “as such” in paragraph 3 of Article 52 – “Paragraph 2 shall exclude the patentability of the subject-matter or activities referred to therein only to the extent to which a European patent application or European patent relates to such subject-matter or activities as such”. In particular, this means that a computer program in its literary form of source code cannot be patented. However, we already established that a computer program has a dual nature. Indeed, on one hand it can be seen as a sequence of textual instructions in source code but it also produces a functionality which can be technical and which can produce an impact in the material world. Therefore, computer programs are not excluded from patentability under all circumstances and the restriction established in Article 52(2)c of the European Patent Convention is not absolute.
The above position is explicitly confirmed also by the Boards of Appeal to the European Patent Office in its decision on case T 0935/97 (Computer program product II/IBM) of 4 February 1999. The Board postulated in its decision that the combined interpretation of the two provisions of the EPC – Article 52(2) and (3) demonstrates that the intention of the legislators was not to exclude from patentability all computer programs. In this sense, the fact that only patent applications relating to software as such are excluded from patentability under the Convention, would mean that patentability may be allowed for patent applications relating to computer programs where the latter are not considered to be computer programs as such. The Board then analyzed the exact meaning of the expression “as such” and clarified that the technical character of an invention is an essential requirement for its patentability. The exclusion from patentability for computer programs as such was thus based on the fact that they are considered to be mere abstract creations that lack technical character. Per argumentum a contrario, computer programs will be considered as patentable inventions when they have a technical character. Thus, we can distinguish between: i) computer programs in the form of mere instructions (not patentable) and ii) computer programs which produce a technical effect (patentable).
This conclusion of the Board raises another difficult question – when exactly computer programs have this required technical character. In its analysis on case T 0935/97 the Board further clarifies that the physical modifications of the computer hardware (the regular flow of electrical currents) that are caused by the execution of the instructions given by the software do not constitute such technical character. This is understandable as such physical modifications are a regular feature of all programs that are run on a computer. In order for a computer program to be patentable it must be capable of bringing about, when running on or loaded into a computer, a further technical effect going beyond the “normal” physical interactions between the program (software) and the computer (hardware) on which it is run. A clear example of such further technical effect would be a computer program that controls the operation of a technical apparatus or an industrial process. In a similar way, computer programs that control the internal functioning or operation of a computer, such as processor load balancing or memory allocation software, normally would produce a further technical effect. Computer programs that determine the emissions by an X-ray device, control the braking system in a self-driving car, compress video or restore distorted images, encrypt/decrypt or sign electronic communications are also examples of software that clearly brings a further technical effect. Computer programs for processing code at low level, such as builders or compilers, may also have a technical character.
Therefore, the general principle under the European Patent Convention is that patents are not granted for software as such but can be granted for software-related inventions where a technical effect can be demonstrated. The term “software-related invention” denotes an invention whose implementation involves the use of a computer, computer network or other programmable apparatus, whereby the invention has one or more features which are realized by means of a computer program. When assessment is made whether the software brings a technical effect, the entire software-related invention as a whole should be considered since the question is whether the whole invention (the software and the device it controls) is patentable. Finally, it must be noted that many software-related inventions which possess technical character, might still fail the other criteria for patentability – novelty, inventive step and industrial applicability, which we will review in the next part of our study on the patent protection of computer programs.