Software patents: setting the record straight
That was a really interesting couple of days at FWIC 2016. Lots of excellent speakers, giving insights into various technical and commercial issues.
Kate Stone of Novalia definitely gave the most engaging presentation!
As a patent attorney, I have to comment on one thing in particular. I heard a number of delegates and even one of the speakers say that software is not patentable in Europe. That is simply not correct: we obtain patents in Europe, and elsewhere, with claims to a computer program (for doing something) all the time.
To elaborate slightly, not all software is patentable, and never has been. But this is generally true: not everything is patentable. As straightforward examples, you cannot patent a piece of art, or a book, or a theme or story for a book, say. These are regarded as “non-technical”. Likewise, you cannot patent a pure business method, again in essence because they are regarded as “non-technical” and for policy reasons. You also cannot patent “scientific theories” or “mathematical methods”, again essentially for policy reasons: no-one should be allowed to patent what is already “out there”, waiting for humans to uncover it.
Software can fall either side of this “technical” line. The European Patent Office and the courts in Europe generally, including in the UK, have developed rules or guides as to what type of software is and isn’t patentable. Similar rules are applied in almost all countries. The rules are fairly clear (though admittedly can sometimes be difficult to apply in some cases). But, putting it simply, the courts and Patent Offices are looking for the software to achieve some “technical effect” in order for the software to be patentable. That “technical effect” might be for example better control of some machine, or a better way of handling large amounts of data, or some improved operation of internal parts of a computer (the way RAM is accessed, or the way a processor handles multiple threads, or some clever DLLs or use of DLLs say). This might be contrasted say with software for a new user interface: if that just makes it “pretty” and has no technical effect on the way the computer operates say, then that will not be patentable.
In my simplistic way, I tend to characterise this by saying that “fluffy” software is not patentable but “hard” software is!
And this, in my view, is right. The whole purpose of the patent system is to encourage innovation, which is good for everyone. To encourage innovation, companies must know that they can invest in a project knowing that they should be able to protect the results and prevent third parties riding off the back of their research and development, with its associated costs. If you are paying your software developers to crack some difficult (technical) problems, the fruits of their work should be protectable.
And on top of whether or not the software is “technical”, it still has to be new and inventive (non-obvious) compared to what is already known in order for it to be patentable, as for all inventions.
I think it fair to say that the position in Europe and the UK on this has been settled and constant for a good number of years. The other extreme is the US. There, they have very recently shifted from a very liberal position a few years ago, where it seemed that practically anything was patentable, to a much more hard-nosed approach, where patents are being invalidated and patent applications refused much more readily at present. (This applies across many technical fields, and not just software.) Unfortunately, the US seems to have lurched from one extreme to another. Hopefully it will settle down in the middle ground soon.
And don’t get me wrong: we have all seen some fairly ludicrous patents that should never have been granted, for all sorts of things. But these are a handful out of the hundreds of thousands of patents which are granted worldwide each year. Just because a system isn’t perfect doesn’t mean it is broken!