The recent decision by the Court of Appeal in Symbian Ltd vs. Comptroller General of Patents (the UK Intellectual Property Office, or UKIPO, to you and me) confirms that software patents are alive and well in the UK.
The decision has brought the approach in the UK more in line with that of the European Patent Office (EPO) and, it is hoped, provided a greater degree of predictability in the patenting of software inventions. Symbian invented software which led to a “better” computer. The Court of Appeal has said that that was patentable.
While software has always been patentable to some degree, the recent practice of the UKIPO has made it difficult to get patent protection for certain types of software invention. This has become particularly noticeable in view of the more lenient approach of the EPO which also grants patents covering the UK.
The commonly mis-held belief that software is not patentable can be traced to the wording used in the UK Patents Act, which explicitly excludes certain inventions from patentability, even if the usual conditions of novelty and inventiveness are satisfied.
The exclusions include mathematical methods and computer programs. However, these are only excluded if the patent relates to that material as such. Similar wording can be found in the European Patent Convention, and UK law is supposed to have essentially the same scope.
Unfortunately, the two patent offices have taken somewhat different approaches to interpreting this limitation when considering whether to grant a patent.
When deciding whether or not an invention is patentable, the EPO has developed tests which look at whether or not the invention has ‘technical character’ or ‘technical effect’.
By any objective standard, just about anything to do with the workings of a computer can reasonably be considered to be technical. Therefore, the logic has been developed to consider the contribution being made by the invention to be patented (in other words, the clever bit).
If the clever bit is just one of the things listed in the exclusions, it will not be patentable. But if it is something more, it will be patentable. For example, an inventor may be trying to patent software which simply computerises a known process, e.g. driving a train.
All that is new is the computer program, which therefore is not patentable. On the other hand, if the computer program has a new technical effect, e.g. it makes the train go faster, it is patentable.
This is the approach validated by the Court of Appeal in Symbian, which has acknowledged that contrary to the earlier view of the UKIPO, software which makes a computer go faster can be patented. In doing so it gave some clear guidance as to when the UKIPO and courts should follow the decisions of the EPO boards of appeal. This in turn should help to avoid different practices developing in the future.
What does all this mean in practical terms?
To answer this, we can consider two cases. The first is my first stumbling steps in writing a computer program to calculate the standard deviation of a set of data (in FORTRAN on punched cards, no less).
This is a computer program (clearly); there is a whole lot that is ‘technical’ going on once loaded and run; BUT all I was doing was using a computer to perform a standard mathematical manipulation. This, any contribution falls wholly within an excluded category: mathematical methods. Merely doing it on a computer did nothing to change this.
The second case, the Symbian case discussed by the Court of Appeal, relates to the way in which functionality is added to a DLL[?] with ordinal linking. Again, this is a computer program and again there is a lot that is ‘technical’ going on when it runs.
However, in this case, the clever bit was to do it in a way that avoided giving the same ordinal link to two different functions and so avoided conflict in the operation of the computer. This clearly has a ‘technical’ effect. As the Court of Appeal said, it is a ‘better’ computer as a result of it (unlike the computer running my program).
So where does this leave us?
The Court of Appeal has clearly said that the UKIPO has been using the wrong standards in deciding which software patents are allowable. It has pushed the UKIPO and courts towards the approach taken by the EPO. Finally, it has confirmed that software is broadly patentable (as if that was in any doubt).
Interestingly, the president of the EPO has recently asked the Enlarged Board of Appeal to look at a number of questions to address perceived inconsistencies in the approach of the EPO’s Boards of Appeal in assessing the patentability of software.
While the EBA will probably refuse to look at these questions, this shows that this subject will continue to exercise minds for some time to come.
Martin Hyden is a European patent attorney with considerable experience of patent prosecution and portfolio management, both private practice and in-house.