As I explained in my report of the recent decision, I believe that there are a number of questionable aspects of the court’s reasoning which leave the law regarding the patentability of computer-implemented business processes in an unsatisfactory state of uncertainty. I am therefore very pleased to see that RA is seeking to take this issue to the next stage, and I am hopeful that the Full Court will agree to hear the appeal.
In many disputes that come before the Federal Court, there is an automatic right of appeal to a Full Bench (three judges) of the court. However, this is not the case where the initial decision is itself the result of an appeal from a decision of the Commissioner of Patents. Subsection 158(2) of the Patents Act 1990 states that:
Except with the leave of the Federal Court, an appeal does not lie to the Full Court of the Federal Court against a judgment or order of a single judge of the Federal Court in the exercise of its jurisdiction to hear and determine appeals from decisions or directions of the Commissioner.
Open Question of ‘Business Method’ PatentabilityThe RA situation has much in common with the case of Grant v Commissioner of Patents, which was the previous significant example in which patent claims relating to a ‘business method’ rejected by the Patent Office on grounds of lack of ‘manner of manufacture’ (i.e. patent-ineligible subject matter) were taken on appeal to the Federal Court. Research Affiliates, like Grant, deals with subject matter on the hazy boundary of patentability, as currently understood. And in both cases, the court upheld the Patent Office decision at first instance, but arrived at its conclusion by way of different reasoning.
The Full Court in Grant accepted an application for leave to appeal, and then (in Grant v Commissioner of Patents  FCAFC 120) upheld the rejection of the claims, based on yet another line of reasoning. While no court opinion ever pleases all of the people all of the time, the Full Court’s decision in Grant was a relatively good one, which certainly clarified some of the main factors which will make an invention ineligible for patent protection. However, Mr Grant’s patent claims were directed to a ‘pure’ business method, in that they did not expressly recite the use of any technology, such as a computer, for performing any of the steps in the method.
As a result, the Full Court decision in Grant left open the question of exactly what additional ‘content’ might be required in an invention to confer patent-eligibility. In particular, it is already clear that computer-implemented inventions can be patentable under Australian law, e.g. the curve calculator in IBM v Commissioner of Patents  FCA 625, the Chinese character generator in CCOM Pty Ltd v Jiejing Pty Ltd  FCA 1168 or the customer loyalty card system in Welcome Real-Time SA v Catuity Inc  FCA 445. However, what is not clear is the extent to which employing a programmed computer in the implementation of an invention might be capable of conferring patent-eligibility upon what might otherwise have been an unpatentable scheme or abstract idea.
In some respects, this question lies at the heart of Research Affiliates. In Grant, which related more to manipulations of legal affairs, rather than purely financial affairs, the Full Court said:
The practice of the law requires, amongst other things, ingenuity and imagination which may produce new kinds of transactions or litigation arguments which could well warrant the description of discoveries. But they are not inventions. Legal advices, schemes, arguments and the like are not a manner of manufacture.
It is not difficult to imagine that a similar statement could quite reasonably be made about the practice of providing investment advice, which might weigh against patentability of the claims in Research Affiliates.
On the other hand, a computer-automated method of constructing passive portfolios and indexes for use in securities trading does not, in itself, constitute the provision of advice. Rather, it may be seen as a tool which can assist an investment advisor in developing products which may be suitable for recommendation to clients. In this sense, the Research Affiliates case may be more akin to Welcome Real-Time, in which the court found a ‘business method’ invention patentable because:
…the Patent is for a method and a device, involving components such as smart cards and POS terminals, in a business; and not just one business but an infinite range of retail businesses.
The court’s recent decision in Research Affiliates did not address the issue of whether this is a case ‘more like’ Grant or Welcome Real-Time. This is something about which users of the patent system – including applicants in this field of technology, and their professional advisors – require greater clarity. As matters stand, the decision does not provide us with any clarity.
Good Prospects of Appeal Being AcceptedIn my view, there are good prospects that the Full Court will accept RA’s application for leave to appeal. While the grounds of appeal are not currently publicly available, it seems to me that there are at least three points on which the court may have fallen into legal error, and which would therefore provide possible basis for the appeal.
- The court focussed on what the RA method produced, calling the index resulting from the method ‘nothing more than a set of data’, ‘simply information’, ‘a set of numbers’ which is ‘no more a manner of manufacture than a bank balance’. The correct inquiry, in my opinion, should be directed to the claimed method itself, not to its product. There is no requirement in any other area of technology that the product of a patentable method must itself satisfy all – or any – of the usual conditions of patentability. The product of a method need not be new, it need not be inventive, and presumably need not be a manner of manufacture. Of course, if it is all of these things, then the product itself is also patentable, independently of the method used to produce it.
- The court was further concerned that RA’s patent specification provides insufficient detail regarding the implementation of the claimed method. Whether or not this is so, in my opinion this is an issue that is properly addressed under section 40 of the Patents Act 1990, and which has nothing to do with ‘manner of manufacture’.
- The court also contended that RA’s case depended upon the proposition ‘that
information of economic significance, once entered into or produced by means of a computer, becomes an economically valuable artificially created state of affairs, and thus patentable,’ and then rejected this proposition. In my view, this is not a valid characterisation of the issues in the case. The question is not whether ‘information’ is made patentable by entry into, or production by, a computer – the court is almost certainly correct that it is not. The question is whether the claimed method, by which the information is produced, is patentable. This is not at all the same thing.
ConclusionI, for one, have my fingers crossed that the Full Court takes on the appeal, and produces a unanimous judgment which leaves the law in a much clearer and more consistent state than it is at present.
While clarity and consistency are most important, to provide all of us with greater certainty, I also hope that the Full Court finds in favour of Research Affiliates. I am on record saying that the original Patent Office decision was ‘clearly wrong’, so it would be nice to be vindicated.
I am a bit of a fan of the Full Court in this regard. Another occasion on which I made an early call was when Apple sought its original interlocutory injunction against Samsung. At the outset I predicted, on the basis of the then-existing case law on such injunctions in patent cases, that Apple would not get an injunction. Justice Bennett subsequently made a liar of me. However, as we all know, the Full Court went on to rule that Justice Bennett got it wrong, and the High Court saw no reason to take the matter any further. I am not ashamed to say that I would love to see a similar outcome here!