Patentability of software and computer-related inventions
This is the first of a two-part post about patentability of software and computer-related inventions in Australia.
This part provides a brief overview of the relevant law and some seminal court decisions to consider when determining whether these types of inventions are inherently protectable. This is by no means a comprehensive review of the case law in this area and there are some points of law and relevant considerations that I have omitted for the sake of brevity. The idea is to provide you with a flavour of some of the more pertinent issues.
Manner of manufacture
Before reviewing the cases, let’s start with some basics. Australian patent legislation, the Patents Act 1990 (Cth), does not specifically exclude computer software from being patented. The Act, however, provides that an invention is only patentable if it constitutes a ‘manner of manufacture’. If you are wondering what on earth this means then congratulations – you are normal! The phrase is derived from an antiquated and long-since-repealed English statute and the Australian courts have, mercifully, confirmed that its literal meaning is not to be interpreted directly. Instead, the broad principles that have been developed by the courts for its application should be considered.
Applying these principles, an invention that consists solely of an abstract idea or mere intellectual information (for example, a pure business scheme) is not a manner of manufacture. It follows that an invention that consists solely of excluded subject matter that is embodied and implemented in a computer program is also excluded. A computer-implemented invention that does not fall squarely within excluded subject matter may, however, be patentable.
NRDC v Commissioner of Patents (1959) 102 CLR 252
The first case to consider is an oldie but a goodie. Whilst it is not a software-related case, NRDC established several important legal principles that have been applied subsequently by the courts in software cases. It was held that to be patentable an invention must (i) entail an end result that is an artificially created state of affairs, (ii) be of utility in a field of economic endeavour and (iii) offer some material advantage in the sense that it belongs to the useful arts (c.f. the fine arts).
If you are thinking this all sounds like something delivered from an ivory tower then you may be forgiven. Admittedly, it is not immediately clear how these principles apply in a software context. Subsequent court decisions have, however, provided some better guidance.
IBM v Commissioner of Patents (1991) 33 FCR 218
OK, time for some tech. In IBM, an improved method for generating curved images using a computer was held to be a patentable manner of manufacture. The method involved an elegant algorithm that used the hardware of a computer that performs integer-based arithmetic to create curved images. Previous methods used floating point computing hardware to create curves which was significantly less computationally efficient.
The Court held that the new method caused a computer’s hardware to operate in a new way foreign to its normal use. The method occasioned a commercially useful result and was more than just an abstract algorithm running on a computer.
CCOM Pty Ltd v Jiejing Pty Ltd (1994) 51 FCR 260
In CCOM, a computer-implemented method for characterising key strokes that enabled Chinese characters to be selected was held to be patentable. The Court made significant reference to the above-discussed principles in NRDC when reaching its decision. In a nutshell, the Court held that the invention achieved an end result that was an artificially created state of affairs (being the retrieval of graphic representations of desired characters). The Court also held that the invention had utility in a field of economic endeavour (being the use of word processors to assemble text in Chinese characters).
Personally, from a practical perspective I do not think that the reasoning given in this case is particularly useful when evaluating patentability of software inventions. Nevertheless, the case at least serves as a good example of an invention that was held to be patentable under the judicial approach that was followed at the time.
Welcome v Catuity (2001) 113 FCR 110
In Catuity, an invention that involved the use of smart cards (an example of an embedded computing technology) to implement multiple trader loyalty programs was also held to be a manner of manufacture. The Court held that the invention comprised a method and device involving tangible components (including smart card devices) used in business that produced an artificial state of affairs (namely, availing multiple loyalty programs on a single smart card). The Court also held that the invention was beneficial in a field of economic endeavour (namely, retail trading).
Grant v Commissioner of Patents (2006) 154 FCR 62
In contrast to the above cases, in Grant a method for protecting assets from unsecured judgment creditors was held to be a mere business method and, therefore, not a patentable manner of manufacture. The Court elaborated on the meaning of artificially created state of affairs, as first articulated in NRDC. The Court held that to be patentable the invention must produce a concrete effect, phenomenon, manifestation or transformation.
Research Affiliates LLC v Commissioner of Patents [2014] FCAFC 150
In Research Affiliates, the full Federal Court of Australia was tasked with deciding whether or not a computer-implemented method for generating an index for use in securities trading was patentable. Although the patent’s claims provided that the method was computer executed, they did not define any specific computing hardware or technologies used or their mode of operation.
The invention’s value in economic endeavour was not disputed. The principal issue was whether the fact that an abstract method required a computer to operate was sufficient to make it a manner of manufacture. The applicant, Research Affiliates, contended that this was sufficient and that its invention, when implemented, employed unique functions or computer instructions to create an artificially created state of affairs. The respondent, the Commissioner of Patents, contended that this was not sufficient and that the invention comprised mere intellectual information.
The Court, as part of its analysis, emphasised that a distinction exists between a pure business method and one that, in practice, results in a new machine or an old machine giving a new result. The Court ultimately held that the applicant’s invention was not a manner of manufacture because it was an abstract method and a computer, as claimed, constituted a mere means, and not the exclusive means, for implementing the method. In the Court’s view, the invention was principally directed to the index generated by the claimed method and this did not occasion the requisite artificial effect.
Commissioner of Patents v RPL Central Pty Ltd [2015] FCAFC 177
In RPL Central, it was held that a computer-implemented method for assessing the competency or qualifications of individuals with respect to recognised standards was not a manner of manufacture. The patent’s two independent claims defined a method and system, respectively, for gathering evidence for the purpose of assessing an individual’s competency relative to a recognized standard.
Both claims recited the specific computing hardware, technologies and data sets used by the invention and the data flows and interaction between these elements that occurred during execution. Despite this, the Court found that the invention entailed a mere business scheme and that its implementation used generic computing hardware operating in a standard way.
Conclusion
It is clear from the above cases that mere computer implementation of an unpatentable abstract method is not, on its own, sufficient to render the method patentable. It is also clear that if the computer implementation only involves generic and routine use of known computing hardware then this is an important factor that must be taken into account when deciding whether or not the invention is a patentable manner of manufacture.
So if an invention of this nature does merely entail such generic and routine use is it necessarily always going to be unpatentable? In a large number of cases the answer will be yes. However, the Courts have emphasised that the manner of manufacture requirement is, ultimately, a threshold requirement (see the NRDC and Grant cases in particular) and that all relevant principles articulated in the applicable case law must be considered.
Part two of this post (coming soon!) will provide some analysis about what generic and routine use means in practice.
This part provides a brief overview of the relevant law and some seminal court decisions to consider when determining whether these types of inventions are inherently protectable. This is by no means a comprehensive review of the case law in this area and there are some points of law and relevant considerations that I have omitted for the sake of brevity. The idea is to provide you with a flavour of some of the more pertinent issues.
Manner of manufacture
Before reviewing the cases, let’s start with some basics. Australian patent legislation, the Patents Act 1990 (Cth), does not specifically exclude computer software from being patented. The Act, however, provides that an invention is only patentable if it constitutes a ‘manner of manufacture’. If you are wondering what on earth this means then congratulations – you are normal! The phrase is derived from an antiquated and long-since-repealed English statute and the Australian courts have, mercifully, confirmed that its literal meaning is not to be interpreted directly. Instead, the broad principles that have been developed by the courts for its application should be considered.
Applying these principles, an invention that consists solely of an abstract idea or mere intellectual information (for example, a pure business scheme) is not a manner of manufacture. It follows that an invention that consists solely of excluded subject matter that is embodied and implemented in a computer program is also excluded. A computer-implemented invention that does not fall squarely within excluded subject matter may, however, be patentable.
NRDC v Commissioner of Patents (1959) 102 CLR 252
The first case to consider is an oldie but a goodie. Whilst it is not a software-related case, NRDC established several important legal principles that have been applied subsequently by the courts in software cases. It was held that to be patentable an invention must (i) entail an end result that is an artificially created state of affairs, (ii) be of utility in a field of economic endeavour and (iii) offer some material advantage in the sense that it belongs to the useful arts (c.f. the fine arts).
If you are thinking this all sounds like something delivered from an ivory tower then you may be forgiven. Admittedly, it is not immediately clear how these principles apply in a software context. Subsequent court decisions have, however, provided some better guidance.
IBM v Commissioner of Patents (1991) 33 FCR 218
OK, time for some tech. In IBM, an improved method for generating curved images using a computer was held to be a patentable manner of manufacture. The method involved an elegant algorithm that used the hardware of a computer that performs integer-based arithmetic to create curved images. Previous methods used floating point computing hardware to create curves which was significantly less computationally efficient.
The Court held that the new method caused a computer’s hardware to operate in a new way foreign to its normal use. The method occasioned a commercially useful result and was more than just an abstract algorithm running on a computer.
CCOM Pty Ltd v Jiejing Pty Ltd (1994) 51 FCR 260
In CCOM, a computer-implemented method for characterising key strokes that enabled Chinese characters to be selected was held to be patentable. The Court made significant reference to the above-discussed principles in NRDC when reaching its decision. In a nutshell, the Court held that the invention achieved an end result that was an artificially created state of affairs (being the retrieval of graphic representations of desired characters). The Court also held that the invention had utility in a field of economic endeavour (being the use of word processors to assemble text in Chinese characters).
Personally, from a practical perspective I do not think that the reasoning given in this case is particularly useful when evaluating patentability of software inventions. Nevertheless, the case at least serves as a good example of an invention that was held to be patentable under the judicial approach that was followed at the time.
Welcome v Catuity (2001) 113 FCR 110
In Catuity, an invention that involved the use of smart cards (an example of an embedded computing technology) to implement multiple trader loyalty programs was also held to be a manner of manufacture. The Court held that the invention comprised a method and device involving tangible components (including smart card devices) used in business that produced an artificial state of affairs (namely, availing multiple loyalty programs on a single smart card). The Court also held that the invention was beneficial in a field of economic endeavour (namely, retail trading).
Grant v Commissioner of Patents (2006) 154 FCR 62
In contrast to the above cases, in Grant a method for protecting assets from unsecured judgment creditors was held to be a mere business method and, therefore, not a patentable manner of manufacture. The Court elaborated on the meaning of artificially created state of affairs, as first articulated in NRDC. The Court held that to be patentable the invention must produce a concrete effect, phenomenon, manifestation or transformation.
Research Affiliates LLC v Commissioner of Patents [2014] FCAFC 150
In Research Affiliates, the full Federal Court of Australia was tasked with deciding whether or not a computer-implemented method for generating an index for use in securities trading was patentable. Although the patent’s claims provided that the method was computer executed, they did not define any specific computing hardware or technologies used or their mode of operation.
The invention’s value in economic endeavour was not disputed. The principal issue was whether the fact that an abstract method required a computer to operate was sufficient to make it a manner of manufacture. The applicant, Research Affiliates, contended that this was sufficient and that its invention, when implemented, employed unique functions or computer instructions to create an artificially created state of affairs. The respondent, the Commissioner of Patents, contended that this was not sufficient and that the invention comprised mere intellectual information.
The Court, as part of its analysis, emphasised that a distinction exists between a pure business method and one that, in practice, results in a new machine or an old machine giving a new result. The Court ultimately held that the applicant’s invention was not a manner of manufacture because it was an abstract method and a computer, as claimed, constituted a mere means, and not the exclusive means, for implementing the method. In the Court’s view, the invention was principally directed to the index generated by the claimed method and this did not occasion the requisite artificial effect.
Commissioner of Patents v RPL Central Pty Ltd [2015] FCAFC 177
In RPL Central, it was held that a computer-implemented method for assessing the competency or qualifications of individuals with respect to recognised standards was not a manner of manufacture. The patent’s two independent claims defined a method and system, respectively, for gathering evidence for the purpose of assessing an individual’s competency relative to a recognized standard.
Both claims recited the specific computing hardware, technologies and data sets used by the invention and the data flows and interaction between these elements that occurred during execution. Despite this, the Court found that the invention entailed a mere business scheme and that its implementation used generic computing hardware operating in a standard way.
Conclusion
It is clear from the above cases that mere computer implementation of an unpatentable abstract method is not, on its own, sufficient to render the method patentable. It is also clear that if the computer implementation only involves generic and routine use of known computing hardware then this is an important factor that must be taken into account when deciding whether or not the invention is a patentable manner of manufacture.
So if an invention of this nature does merely entail such generic and routine use is it necessarily always going to be unpatentable? In a large number of cases the answer will be yes. However, the Courts have emphasised that the manner of manufacture requirement is, ultimately, a threshold requirement (see the NRDC and Grant cases in particular) and that all relevant principles articulated in the applicable case law must be considered.
Part two of this post (coming soon!) will provide some analysis about what generic and routine use means in practice.