Key Points:

While Australia appears to be following the US lead, the recent decision in Research Affiliates is unlikely to be the last word on the patentability of computer-implemented methods in Australia.

In October last year, we analysed two decisions of the Federal Court of Australia on the patentability of computer-implemented methods. Those decisions appeared to be somewhat contradictory, leaving patent owners in the dark as to whether patents in this area are likely to be valid.

Earlier this year, the US Supreme Court handed down its decision in Alice Corporation Pty Ltd v CLS Bank International 134 S Ct 2347 (2014), holding that the patent in issue was invalid because the claims were directed to an abstract idea of intermediated settlement, which was "a fundamental economic practice long prevalent in our system of commerce". Even though all of the claims were implemented using a computer, the Court found that the claims merely required generic computer implementation, which was insufficient to transform that "abstract idea" into a patent eligible invention.

Meanwhile in Australia, the Full Federal Court has just handed down its appeal decision in Research Affiliates LLC v Commissioner of Patents [2014] FCAFC 150, which considered similar issues. Despite this, just where the line is to be drawn in these cases is by no means clear. In this article, we try to dispel some of the confusion.

Requirements for validity

To be patentable in Australia, a claimed invention must be "a manner of manufacture" according to the developed case law. According to the High Court's judgment in NRDC, that will generally be the case if the invention:

  • involves human intervention, in the sense that the invention involves a mode or manner of achieving an end result which is artificially created; and
  • has industrial applicability, in the sense that it relates to a state of affairs of utility in afield of economic endeavour.

As a result, none of the following are patentable subject matter:

  • intellectual information in and of itself which does not have some kind of artificially created physical effect (for example, a mathematical formula, abstract idea, scientific principle or method of calculation);
  • an invention for the betterment of the fine arts, as opposed to the useful arts; or
  • a mere plan or a scheme (such as a plan for the efficient conduct of a business or pure "business method").

That's all fine, but what is the effect of plonking a computer in the middle?

Not much, the Full Court and US Supreme Court think

The patent in Research Affiliates claimed a method of constructing a weighted index and portfolio of securities: in other words, a clever way to pick and choose which stocks to pop in the super fund. It was able to do so not only on the basis of standard variables (eg. market capitalisation) but on the basis of non-financial or non-capitalisation metrics (eg. companies with CEOs that attended a particular university). In each case, a computer is used to receive the data and provide "weighting means" to produce the index.

The trial judge found that the patent was invalid because, among other reasons, the invention did not produce the requisite "artificially created state of affairs": it was essentially a business scheme which could have been performed without the use of a computer (eg. with a pen and paper – although it would obviously take a lot longer).

On appeal, the patentee (RA) argued that the trial judge was wrong because both the index that is produced, and the method to produce the index, result in physical effects in a computer, which is an artificially created state of affairs. In other words, argued RA, the fact that the method is implemented in a computer satisfies the requirement in NRDC for an artificially created state of affairs.

By contrast, the Commissioner of Patents (which rejected the patent in the first instance and was defending its decision to do so) argued that the result of the claimed method is an index and an index is not the kind of thing that can be patentable. While RA's invention may be clever and may represent a better method of constructing an index intellectually, said the Commissioner, it has no effect on "the computer architecture". In other words, merely to implement an unpatentable method in a computer does not convert what is otherwise inherently unpatentable into something that is patentable.

The Full Court recognised that the use of a computer necessarily involves the writing of information into the computer’s memory. This means that there are a number of "physical effects" in the sense of transformed data and memory storage during the process. Despite this, the Full Court sided with the Commissioner, finding that the "inexorable conclusion applying the principles of patentability, is that the present claim is not to a patentable method". The effect of RA's argument, the Full Court said, would be that the mere implementation of any abstract idea or scheme in a well-known machine (in this case, a computer) would be sufficient to render that unpatentable subject matter patentable because it gives rise to an "artificial effect". This approach, the Full Court said, would be inconsistent with NRDC and would elevate form over substance.

However, to what extent will the involvement of a computer render an otherwise patent ineligible concept patentable?

Where is the line?

In answering this question, the Full Court in Research Affiliates tell us that there is no formula to be mechanically applied. It is a question of understanding what has been the work of, the output of, and the result of, human ingenuity in the claimed invention, and then applying the NRDC principles.

One possible means to carry out this analysis is to ask whether the invention could be performed without the use of a computer and whether, if so, that non-computer implemented invention would be patentable. This is essentially the approach taken in the US, where the courts say that the use of a computer in an otherwise patent-ineligible process for no more than its most basic function – making calculations or computations – fails to circumvent the prohibition against patenting abstract ideas and mental processes. Similarly, in Australia and in the United Kingdom, whether the involvement of the computer is "foreign to the normal use of computers" has been considered relevant.

From another perspective, we can ask whether the invention involves an advancement in computers, or the use of computers, over and above what might otherwise appear to be unpatentable. In Alice, Justice Thomas explained the US position by stating that there is a distinction between mere implementation of an abstract idea in a computer and implementation of an abstract idea in a computer that creates an improvement in the computer. Similarly in Australia, as explained in Research Affiliates, a distinction has been drawn between a technological innovation (which is patentable), and a business innovation (which is not).

Following on from this concept, and although the Australian courts may not like to say so expressly, it seems that the way in which the relevant patent has been drafted as a whole – not just the claims defining the monopoly – will be a significant factor in determining whether a patent in this area will be valid. In this regard, Australian courts have considered the extent to which the computer is "inextricably linked" to the invention, which will often be determined by considering the extent to which the specification explains the way in which the computer is to be used in the method. For instance, the Full Court in Research Affiliates said:

"Aside from the method being one that is ‘computer-implemented’, there is nothing in the character of the steps comprising the generation of the securities portfolio index that relates the method to any particular hardware or software implementation. There is no description in the specification providing the detail of computer implementation." [emphasis added]

Similarly, the Full Court noted:

"Although the summary of the invention and the exemplary embodiments do make reference to computers, it is to be noted that what is missing from the title of the invention, the described field of the invention and the detailed description of the invention is any reference to a computer, even though the claims limit the method of the invention to one that is computer-implemented." [emphasis added]

A final note – is Alice all she's made out to be?

The Full Court noted that the patent in Research Affiliates would also be invalid in the US, on the basis of the principles discussed in Alice. Although some say the policy rationale behind the "abstract idea" prohibition in the US is the same as Australia's "manner of manufacture" requirement, there are important differences in the Australian and US patent systems which should give Australian courts pause before placing too much reliance on the analyses of US courts, not only on this question, but more generally.

Further, the Alice decision has been widely criticised, with many arguing that the analyses of the US Supreme Court may be helpful on the question of whether an invention is obvious, but should not be used to close the door to patentability entirely.

Meanwhile, making its way through the courts in Australia is the appeal in the other case we discussed last year – the RPL Central case. The appeal in that case was put on hold pending the outcome of the Research Affiliates case. The decision on appeal in RPL Central will assist in helping businesses understand where the line of patentability is drawn in the case of patents involving implementation via a computer. In its analysis of the decision at first instance in RPL Central, the Full Court in Research Affiliates appears to indicate that the invention claimed RPL Central (at least, as analysed by the trial judge in that case) did result in an artificially created state of affairs. As such, the decision in Research Affiliates appears to be a very narrow one – the effect of which is to exclude from patentability only the mere implementation of a method via a computer.

In addition, although the Full Court's decision in Research Affiliates is binding on lower Australian courts for now, it is yet to be seen whether RA will seek leave to appeal to the High Court. As such, this story might not have reached its conclusion quite yet.