Can computer programmes be granted patents? On February 19, India’s patent office wisely answered this question in the negative, putting an end to months of ambiguity over the patentability of computer programmes. In this process, the patent office, called the Office of the Controller General of Patents, Designs & Trade Marks, effectively reversed an August 2015 guideline that had triggered the ambiguity in the first place. Till that guideline came, India’s stance on this issue had been clear through a 2002 amendment to the Patents Act: that software per se was not to qualify for patent protection. However, lawmakers also recognised that the intention must not be to reject inventions involving software that “may include certain other things, ancillary thereto or developed thereon”. Experts have interpreted this exception to refer to innovations in both software and hardware. The 2015 guideline threatened to unsettle that nuance. According to that, technical advancements could be sufficient grounds on which to confer patents. Its nullification is welcome as such rules, though seemingly on the side of innovation, do not enable a level playing ground. For starters, the share of patents held by Indians has traditionally been low, and it continues to be so. Also, the field of software is dominated by corporate giants with deep pockets and significant expertise, and they can easily ‘out-patent’ the others out of business. The smaller companies and start-ups — and there are far too many aspirants with that profile — then not only have to spend huge sums of money to protect their work, but they also have to be financially and operationally ready to defend themselves.
The patent office hasn’t left it at just that. It has also issued a three-stage test to examine applications of computer-related inventions. Step one is to interpret the claim. Once that is done, step two is to deny the claim in case the “contribution lies only in mathematical model, business method or algorithm”. Step three is to assess if the invention is claimed in the field of software in conjunction with a novel hardware. The important point to note is the recognition that software in itself is never patentable. This is a prudent stance, because there are inherent problems in figuring out if software is patentable or not. And this is true the world over. Germany and New Zealand exclude software from patentability. In many other parts of the world, the positions are nuanced, like the one taken by the European Patent Convention, which does not entertain applications when they pertain to computer programmes as such, but it does have an open mind when they lead to “non-obvious” contributions. In the U.S., a more open policy has led to a flood of patents, and consequently the negative connotation that the term ‘patent thicket’ carries now. There is a more important reason for holding back software from a patents regime. And this goes back to what MIT researchers James Bessen and Eric Maskin showed many years ago: imitation promotes innovation. Patents are a hindrance here.