2005-06-14 Briefing Document from Sharon Bowles
Computer Implemented inventions aka Software Patents
This proposed directive is not about what you run on your desktop computer it concerns everything that can be plugged in or switched on, from mobile phones to medical body scanners; from aircraft controls to wristwatches. Nor is it a David versus Goliath battle 70% of Europes patents for computer implemented inventions are owned by SMEs.
Its potential impact is this wide because it covers any programmable apparatus controlled by a computer program. Silicon chip devices, fundamental nowadays to all electronics, are made to work by the varied application of charges (programming), and thus are programmable apparatus. Amendments proposed by the Parliament's rapporteur to change this definition to any invention the performance of which involves the use of programmable apparatus puts coverage of all electronics beyond all doubt.
Definitions are the bugbear of this directive. What it really concerns is a technical method implemented via program instructions and how to enforce patents for such methods. A purchaser of a copied method, rendered in the form of a program, is the direct infringer of a patent for the method. The real culprit, the provider of the program, is not a direct infringer but can be pursued for contributory infringement, that is providing the essential elements of the patented invention. However, that is a more complicated legal action and does not cover importation. Copyright is not sufficient because for any technical method, which is really where the invention lies, there will be numerous ways of actually writing the program.
Patent law excludes computer programs from patentability, with the added proviso that this exclusion only applies to programs as such. Debate on whether this means everything to do with programs is excluded, or just the detail of listings or code (which is protected by copyright), or something in between, has raged. In recent years the European Patent Office reached the pragmatic solution that when something technical was going on (a technical effect) then the program was more than just software as such and it was patentable. When there was no technical effect (e.g. as for business methods) then it was not patentable. Such a pragmatic solution is not out of line, after all an area of law that operates at the forefront of technology has to move with technology. In the 19th century the debate was whether dying cloth constituted an invention capable of industrial application. Some suggest that the EPO has gone beyond its remit because there have been different national responses to the potential validity of their solution. However in the conferences setting up the European Patent Convention the guideline for the EPO was that it should allow a patent when there was a probability of it being held valid in half of the member states. This guideline has not been breached.
The Common Position does not propose to allow patents for all kinds of software whatever its purpose. It proposes to allow a software product claim for an invention when there is also a method or apparatus claim for the invention in the same patent this is the way of tying it to the specific technology disclosed and of course all the usual requirements for patentability of novelty, industrial application and non-obviousness still apply. However, the wide definition of a computer implemented invention (programmable apparatus etc) means all other restrictions put on CIIs will apply not just to the software product claims but potentially to method and apparatus claims of any electronic invention where the use of programmable apparatus inevitably forms a part. Most of the amendments include terms that, imposed on these electronic products and methods, will pull the carpet from under the European electronics industry leaving them unable to protect their vital home market. Other countries would still enjoy home market protection, enforcing patents against European companies. Exactly the same considerations apply against proposals for free interoperability licences in the EU, a notion emanating from US rather than European industry which should say it all! And that does not mean just big companies would be hurt. Activity, in terms patents granted by the European Patent Office for computer implemented inventions, is 70% by SMEs. Academic institutions are also active in this field deriving income from patents. The relevant sector employs millions.
Among the most damaging proposed amendments are those that seek to exclude data processing or information handling. Nobody wants to patent simple data manoeuvres: the cost and timescale is too great, and in any event they are not patentable if they are trivial and obvious. But complex signal processing that is the backbone of electronics is also data processing because digital signals are data. It would be a very strange thing to allow a patent for an electronic circuit in analogue hardware but not allow the same protection to its digital rendering. Information is also an expression used in radar and electronics for decades before the information age and refers again to signals and their content as compared to background or noise. So information handling is the same as signal handling. Not for nothing did the drafters of patent law choose the information exclusion as presentation of information. If amendments such as these are passed then the directive would have to be dropped.
Apart from the waste of time and fears aroused from scaremongering many professionals could live happily with that. But there are some good amendments that fill holes in the Common Position, in particular some that would strengthen monitoring and for the first time provide a way to influence the European Patent Office, given that it is NOT an EU institution. Such monitoring and influence should be taken very seriously if the EPO is to become the agency for the Community Patent when that finally happens. In practical terms these monitoring amendments making use of the Member States positions on the Administrative Council of the EPO will be far more useful than tweaking definitions of technical effect because it is felt (with some justification, but note the valid in half the member states remit) that in its worst decisions the EPO has rendered the term a little shop soiled. Any new term could receive the same treatment at the EPO.
Are any of the fears from the software lobby justified? Those who have been led to believe that the directive is changing the status quo have nothing to fear because it is not substantially changing the status quo. Those who want to deny patents for any kind of method, product or program involving data manipulation will not succeed because they are turning back the clock on over 50 years of technology and precedent and the price for it, the destruction of the European electronics industry, is too high and the directive must fall instead. Those who fear the cost of licensing products such as Adobe Distiller which creates pdfs and allows rapid transmission of data should put their faith in the fact that patentee companies do not embroil themselves in complex negotiations for every licence, terms have to be reasonable (otherwise competition policy comes into play) and if the common position is amended to include compulsory licensing proposals for de facto standards the directive would improve their position.
