Do you remember the Queen’s Silver Jubilee in 1977? Street parties and parades were set against a second wave of punk rock, and all organised (or not, in the case of punk) without a mobile phone or e-mail in sight, let alone a text message or tweet. Some smaller computers were starting to challenge a landscape dominated by mainframes and minis, but IBM PCs and Macs were still years off. In the same year, the shiny new European Patent Convention came into effect, although the first European patent applications were not filed until June 1978, and the event could scarcely have been noticed by more than a few thousand people across the whole of Europe.
The European Patent Convention (EPC) was an exciting development in many ways, at a time when ‘Europe’ was still a novelty over here, the UK and Ireland having joined the EEC fairly recently in 1973. However, when you realise that the UK Patent Office alone (now the UK IPO) had already published well over a thousand patents relating to computers by this time, it is perhaps not surprising that the EPC explicitly mentioned ‘programs for computers’. What might be surprising is that Article 52 of the EPC, far from heralding in a new age of computer technology with enthusiasm, actually seems to ban them: ‘(2) The following in particular shall not be regarded as inventions… (c) schemes, rules and methods for performing mental acts, playing games or doing business, and programs for computers…‘.
The story for the next 30 years at the European Patent Office (EPO) is one of how to reconcile this apparent ban with the march of technology and with common sense. A later part of the same Article 52 in the EPC provides a slippery toehold: ‘(3) The provisions of paragraph 2 shall exclude patentability of the subject-matter or activities referred to in that provision only to the extent to which a European patent application or European patent relates to such subject-matter or activities as such‘. This, in combination with an otherwise obscure and apparently innocent use of the word ‘technical’ in the EPC Rules, gradually lead to the modern practice of considering whether an invention which is implemented solely as a computer program solves ‘a technical problem’. In this way, the European Patent Office now grants many thousands of patents a year to inventions which are implemented as computer programs. The major national patent offices of Europe, with the notable exception of the UK office, examine patent applications in much the same way.
An interesting aside which demonstrates the intensity of feeling which can be generated by this subject relates to an attempt by the European Commission between 2002 and 2003 to codify and harmonize across Europe the degree to which computer program inventions should be protectable by patents. From the perspective of patent attorneys, the directive proposed by the Commission was largely an innocent restatement of existing law and practice. However, it whipped up a fervour of anti-patent feeling across the software development community, and in August 2003 hundreds of people gathered for a protest rally outside the EU parliament in Brussels, and an enormous raft of proposed parliamentary amendments ranging from the pragmatic to the completely baffling sank the directive without trace. Even now, six years later, I am still met by a degree of polite hostility when talking to some computer programmers in my patent attorney role.
I can only suppose that some of that anger found its way into the UK IPO. In 2003 there was still relatively little case law from the UK courts regarding the interpretation of the same ‘programs for computers‘ and ‘as such‘ conditions which appear in the 1977 UK Patents Act, and what there was seemed sympathetic to the EPO position. Computer-related patent applications were being examined, and granted by UK IPO examiners in a manner broadly in line with the EPO, with the term ‘technical effect’ appearing frequently. Then, in July 2005, a High Court decision refusing two patent applications belonging to an American company, CFPH LLC, contained some worrying indications.
It is perhaps no surprise that Peter Prescott QC, sitting as deputy judge, found against the applications. These claimed computerized wagering systems which I think would also have struggled woefully before the European Patent Office, in line with a broader European view that the Americans were getting far too carried away with easy granting of patents to business schemes and similar that had no real link with ‘technology’. The CFPH judgment acknowledged that the UK IPO did not use quite the same reasoning as the EPO in assessing the patentability of computer program inventions, even though they were working under what is essentially the same legislation, and went on to say that ‘It is fairly obvious that our Patent Office is not very enthusiastic about having to track every twist and turn of the EPO’s reasoning…‘. In some final remarks it is said of business method patents that ‘130. …. An accumulation of patents of that sort…may be a serious barrier to entry. 131. The only safeguard against that wrong – and it is a wrong – is the vigilance of the Patent Office.‘
The CFPH judgment was followed in late 2006 by a UK Appeal Court decision about two patent applications in which the UK IPO urges upon the court a four-step test. The Macrossan patent application, already refused by the UK IPO, lay even deeper in the territory of business methods than the CFPH application mentioned above, whereas the Aerotel application related slightly more favourably to telephone network systems and came out of court intact.
The new four step ‘Aerotel/Macrossan’ test did not, read literally, impose any major differentiation over either the existing UK IPO or EPO approaches. One step of the test asks if the ‘contribution’ to the prior art made by the invention ‘falls solely within the excluded subject matter’, which the EPO would answer with a ‘no’ if a computer was being used. Another step asks if the ‘contribution is actually technical in nature’, an issue which the EPO would consider by asking whether a technical problem was solved in a technical way. It was therefore telling to see the speed and enthusiasm with which UK IPO examiners now imposed an almost complete ban on computer program inventions. On a scale of 1 to 10, where 1 represents granting only the most machinery based computer program patents, and 10 represents granting the most abstract of business methods, the UK IPO had reached somewhere around 2 to 3.
To a UK patent attorney it looked a lot like the UK IPO, rather than simply following the law, had its own agenda of restricting the patentability of computer program inventions, and only needed to persuade the courts to look unfavourably on two or three patent applications for a wholesale change of patent granting policy to be swept into place. This might sound over dramatic, but a July 2009 letter, from the Chartered Institute of Patent Attorneys to the Minister of State for Higher Education and Intellectual Property, expressed a real concern over a mixing together of the Government policy and patent granting roles of the UK IPO.
You can’t really understand the human angle to this story without hearing about the State Street Bank case. The US Patent and Trademark Office (USPTO) had always granted patents to abstract concepts, computer programs and the like, occasionally, and the case law had drifted one way and another for decades. But the rapid increase in the commercial use of the Internet in the 1990s was building up a head of pressure, and that pressure was released by the State Street Bank vs Signature Financial Group decision of the US Court of Appeals of the Federal Circuit (CAFC, which specialises in IP matters) in 1998. The details of that decision are beyond the scope of this brief article, but it was probably this case which opened the eyes of US lawyers and businesses alike to the possibilities and triggered a huge wave of business method patent applications in the US, based largely on the rapidly expanding (and soon to burst) dot-com bubble. And when US patent applications are filed, follow-on applications are soon filed in the rest of the world, including the UK and the EPO. Never mind that the dot-com bubble burst in 2000 – that wave kept on coming, and by 2003 and the CFPH decision mentioned above, the UK IPO examiners and their managers were probably seriously fed up with it.
On my scale of 1 to 10, the USPTO was probably now at 10 – patents were allowed on the full range of software inventions and beyond into really quite abstract territory, and as the dot-com bubble became history, a ground swell of opinion against over liberal patent granting in the US started to have a political effect felt at the USPTO. Urban myths from that time about patents on a method of swinging a swing, and the Amazon one click patent are only partly myth. Then in 2008 another decision of the CAFC considered a patent application by Bernard L. Bilski and Rand Warsaw. The patent application related to commodity trading, largely in the abstract in the sense that no computer was mentioned in the patent claims. The judges displayed an interesting diversity of opinion, perhaps in a range of 6 to 10 on our scale, but the patent application was refused and some new patentability tests, more baffling than the UK’s Macrossan test, and possibly placing US law around 9 on our scale, were introduced. The decision was appealed further and the matter is currently being considered in the US Supreme Court. From the barrage of questions against Bilski’s attorneys at Oral Argument in November 2009, some are concluding that the justices are more inclined to restrict the scope of patentability in the US further, rather than take it back towards 10.
To fill out our scale a little more, one might place engine management systems and microcontrollers in washing machines down at level 1, because the computer and program are directly controlling a mechanism of some kind. Controlling less solid entities such as TV signals and mobile telephone data streams might come in around 2 to 3. In modern engineering, these applications are pure software manipulating pure data, albeit for some distant purpose such as telephony or television, and as such arguably fail the current UK test, although it is doubtful that the UK IPO would dare go so far, as this would imply wholesale destruction of the patent infrastructure behind mobile telephones, digital television and the like.
Image analysis from ultrasound machines and more esoteric aspects of operating systems and software engineering come in around 4. As we move more into ‘applications’ areas such as DNA molecule analysis and drug design, graphical improvement and character control in video games and the like, towards software engineering aspects of ‘applications’ such as word processors, internet search engines, and databases, we move through 5 and 6, and reach the limit of patents allowed by the European Patent Office, and other national patent offices in Europe.
Business management software, more ‘technical’ e-commerce ideas, and complex bank trading software fall around 7 to 8, with patent offices in Korea, Australia and Japan probably not going further than this. Levels 9 and 10 are reserved for almost pure business methods and similar, perhaps implemented on a computer for a level 9, or not for a level 10. The USPTO did grant patent applications in this bracket during the golden decade of 1998 to 2008, but is probably now moving down towards level 8.
So what will happen to the poor UK IPO, isolated and alone at the other end of the scale, abandoned to its own thoughts by applicants in this field who are free to pursue their patents elsewhere? A series of pragmatic UK High Court and appeal court decisions over the past couple of years, including the ‘Symbian’ case relating to dynamic link libraries (from which the UK IPO was refused leave to appeal), have probably pulled the UK IPO up towards a level 4, but still well short of the European Patent Office position. Of course, the UK IPO has the moral support of India, the only other territory broadly in line with the UK position for which I regularly have reason to consider this topic. I understand that India keeps one eye on parallel British developments in this area.
Apart from waiting for the outcome from the Bilski appeal in the US Supreme Court, we are also waiting to hear from the EPO Enlarged Board of Appeal (which provides the highest level of decisions for the European Patent Office) on this general topic. In October 2008 Alison Brimelow, President of the European Patent Office, put a long series of questions to the Enlarged Board asking them to rule on conflicting positions in the EPO case law. Well argued amicus briefs were filed by a wide range of parties, happily with more common sense shown than in the European Directive debacle around 2003, and including a purrfectly sensible one filed (in the name of) a knowledgeable IP cat (http://www.epo.org/patents/appeals/eba-decisions/pending/briefs.html – entry by Sylvestris, T.). Given that the EPO Appeal Boards had already worked so hard to find a now fairly consistent and established approach for considering computer implemented inventions, it does rather seem as if Mrs Brimelow has tried to foist the ‘British Problem’ upon the EPO. We await the outcome with interest.
Jonathan Palmer is a Partner at Boult Wade Tennant. Jonathan is a Chartered and European Patent Attorney in the High Tech and Electrical Group.
