A Policy A Day: Software Patents

In the lead-up to the election, we are examining a policy a day. We're exploring a variety of policy areas, explaining the background and analysing some of the policy options, with a mixture of technocracy and values-based approaches. Inevitably, some opinion will make its way in and we make no apology for that - after all, we're voters too. A list of all the articles is available here. Enjoy!

Today's post is written by Jack Robinson

What is a Patent?
For as long as I can remember, there has been a never ending discussion on intellectual property rights and the domain of the computer. Piracy, privacy, and copyright infringement have been all over the news and media consistently since the internet began, either condemning it or praising it. However, patents have mostly lived an isolated life away from the media spotlight.

Patents are comparable to copyright. Copyright is protections automatically given to an author for an expression of ideas; if I write a book, as soon as it has been written, I have copyright on it. Patents, however, are a set of exclusive rights that a government can grant an inventor for their novel invention in return for publicly disclosing their design in detail. Where copyright covers expressions of ideas, patents cover the idea itself.

On the surface, patents make sense - consider a small entrepreneur who has invented a new type of can opener. The kitchen utensil market is already awash with large manufacturers who have vastly greater resources than the inventor. If they were to enter the market with the same invention, the big companies could potentially undercut the inventor and take their business. By gaining a patent, the inventor can profit off of their invention for a period of time. In return, the government has facilitated the sharing of knowledge via the inventor producing detailed descriptions of their invention. It is important to note that since patents exist on detailed ideas of inventions, the patent-holder is not required to actually build what they have patented. However, they have the legal right to exclude anyone else from manufacturing it for a period of time, generally up to twenty years.

So, it kinda sounds good on paper, doesn’t it? What about for software?

Software Patents
The issue with software is that software is rather abstract. The value it provides is primarily in its intent, and its implementation can be done in any number of ways. If I asked 10 different developers to implement a simple game of Rock Paper Scissors, each would come back with a different implementation - a simple search on Github, a host of many code repositories, lists over 7,000 different implementations and even on the third page I’d wager it is nigh impossible to prove that any are exactly the same in implementation.

Compare this to mechanical inventions, where its intent is wholly embodied by its implementation. If you were reading a patent application for a new type of toaster, you’d be able to verify its intent by its implementation, and use the implementation to further ensure that it is indeed a new invention. With software, you don’t have the fallback option of a physical device to qualify it as a novel invention - no software patent is applied for with concrete code examples. It’s this abstract nature of software that is also its downfall when it comes to the quality of patents.

In 2011, NPR did a piece called ‘When Patents Attack,’ where Laura Sydell went to M-Cam, a company that assesses patents for governments, banks, and businesses, and used a piece of software that trawls patent databases to gauge the quality of a patent. They were assessing the quality of a particular patent, filed in 1993 by a man named Chris Crawford for an ‘online backup system’. What Sydell discovered was that while Chris Crawford was awaiting his patent to be issued (which it was in 1998), over 5,000 other applications were being filed for the same thing. This is a patent for a ‘method and apparatus for efficiently backing up files using multiple computer systems’ awarded to Oracle in 1999, whereas this is a ‘method and apparatus for efficiently mirroring data in a remote data storage system’ also awarded in 1999 to the EMC Corporation. That’s three patents, all issued around the same time for an online backup system similar to Dropbox, yet due to their wording are all valid and upheld patents.

While larger companies battle it out in the courtroom over their own patent infringement (Samsung vs Apple and Google vs Oracle are two recent examples of this), the risk is more with small and medium startup companies without the vast legal resources that companies like Apple have.

In the past, smaller companies would place their patents under the protection of something known as a Non-practicing Entity, or NPE. NPE’s were set up so that these small companies had the ability to defend their patent against infringement on their patents, however, in recent years, these companies have become more known under a different name - Patent Trolls. Practising lawyers can collect large numbers of patents and aggressively pursue any and all that are infringing on their patent, and unfortunately, they’ve gotten rather good at it. In 2012, Bessen and Meurer of Boston University estimated that legal costs directly attributed to cases involving these NPE’s was $29 Billion dollars in the United States, 37% of which came from cases against small and medium businesses.

It seems that these Patent Trolls/NPEs found huge value in these low-quality software patents, as they were able to use shoddy patent descriptions and apply them in large blanket cases against multiple defendants. In 2013, the United States Government Accountability Office released a report that showed the number of defendants in patent cases increased 129% from 2007 to 2011, and that 89% of this increase was in cases involving software. Since these NPE’s are run by lawyers, they can defer a lot of legal fees as they are representing themselves, further increasing the number of intellectual property lawsuits.

How, in any capacity, can small to medium sized businesses protect themselves from such an onslaught of potential patent infringement? As a software developer who one day wants to perhaps start their own company, the idea of patent trolling is scary. How do I know if I’m infringing on shakily written up patents, and how can I protect myself if somebody is coming after me? If I’m an established company, what is the point of investing in further research and development if I’m just going to risk infringing on some dumb patent from 2000? Yes, that is a patent on toast.

New Zealand’s legal system is much less viable for patent trolls to exist, but there is still definite risk for these lower quality patents to affect local software development. The government, acknowledging that this sort of behaviour is detrimental to innovation, decided in 2013 to prevent patents on software.

New Zealand Software Patents
With the passing of the Patents Act 2013, the New Zealand government decreed that, under patent law, software is not* an invention, and therefore cannot be patented. This is with a massive star next to the ‘not’, as there are some exceptions to the law. The law change exists for purely software related inventions, such as algorithms or pieces of functionality; if the software is directly tied to a physical piece of hardware (known as embedded software), then it is potentially patentable [1]. While this isn’t a full 100% banning of software patents, it is definitely a step in the right direction to removing the ability to submit and be issued low-quality patents.

The law change was widely supported by critics, both locally and worldwide, with many different publications covering the news. 81% of members of IT Professionals New Zealand supported the law, stating that “there are a number of reasons against software patents, however, on balance the [members] and the majority of the ICT profession believe they are harmful to innovation of the ICT sector and New Zealand as a whole.”

However, multinational agreements could put all of that at risk. New Zealand is a big fan of free-trade agreements and similar economic partnerships, with currently twelve in force and another ten or so in active discussions. In such agreements, there are generally stipulations that each country needs to come to terms with in order to allow such trade to flow freely - protections on their exports and economy that the signatories need to adhere to. Of recent note, the highly controversial Trans-Pacific Partnership Agreement (TPPA) is an example where lengthy stipulations were requested of differing member nations in order to agree on free trade. In a leak from the discussions, it appeared that Mexico was the only member nation that would hold onto laws against patents on software.

Essentially, it appeared that the United States pressured New Zealand into agreeing to allow software patents again. While Trump has actively denounced the agreement and made movements to retract any United States involvement with the TPPA, the potential precedent set is still a dangerous one – other countries have seen that software patenting is something that we’re willing to compromise on.

In terms of the 2017 New Zealand Election, the TPPA is still in active development. Bill English and National still support the ongoing agreement process with remaining member nations and are looking to get eleven signatories by 2020. Labour has said that they would not pull out of the trade agreement either, but would change aspects of the agreement that are not in line with their policies. The Green party, who were behind the original reforms of the Patent Act, have stated that they are not in agreement with the TPPA, and have consistently stated that they are proud of their work in internet rights and freedoms, copyright, and software patents. It’s Our Future has rated both the Maori Party and NZ First as two parties that have a clear desire to reform the way we have been negotiating the TPPA, whereas TOP has not defined a clear position. To be fair, the TPPA is a much more complicated agreement than just intellectual property, so if a party is for or against it, it is not a clear indicator of support for the reformed Patent Act.

Conclusion
After all that, are patents on software a key discussion point that parties should be having in the upcoming election? Probably not, but that doesn’t mean we should ignore them. The nature of software patents is no doubt a contributing factor in the issuing of low-quality patents, weakening the integrity of the patent system and opening up opportunity for abuse. These risks negatively affect the ability of start-ups and companies to invest in research and development, as they have the potential to infringe on an international patent written decades in the past that could still be used against them. This is why I think software patents are bad for innovation, and this is why I think New Zealand made the right decision in preventing such patents to be made. Of note in this election, however, is that the winning parties could dictate how these policies are treated in the near future.

Jack Robinson is a recent graduate, and software developer currently working in Wellington. He wants to remind the reader that he isn’t a lawyer, as well as he was rather upset that his university login expired right before writing this, meaning he couldn’t sneak in and use the library database to research for this post. He is generally mystified by politics but admits he’s left leaning in many of his opinions.

[1] For a clear example of the difference between pure software and embedded software, I suggest taking a look at this Ministry of Business, Innovation and Employment description.