G 1/19, which admits the patentability of a computer-implemented simulation, was the second opportunity for the Enlarged Board of Appeal to rule on the assessment of the patentability of computer-implemented inventions. Did it take advantage of this One More Chance or was it only One More Time?
At any rate, here, I will only briefly comment these statements in relation to my favorite topic: technicality. And we will see that these statements are interesting! Both on the technicality criterion itself, on the one hand, and on the assessment of technicality, on the other hand.
Technicality criterion
On the technicality criterion, I have some comments on the basis for this criterion and on its definition by the Enlarged Board.
Concerning the basis, the Enlarged Board recalls, as it is now customary inventions must be patentable “in all fields of technology” (Article 27 TRIPS, introduced into article 52 EPC in 2000). Thus, according to the Enlarged Board, this expression would imply a requirement of technicality, which does not appear explicitly in the EPC. I believe this to be wrong for two reasons.
Firstly, the expression “in all fields of technology” was introduced into the TRIPS Agreement thanks the United States in order to widen the field of patentability as much as possible. On the contrary, the technical character criterion limits it. Secondly, technical and technological are two distinct words that are often confused, particularly in English, whereas in reality technology is knowledge about a technique. Knowledge about a technique is not the technique: science can provide knowledge about a technique but is not technical knowledge, for example.
Neither the TRIPS nor the European Patent Convention contain an adequate basis for the requirement of technical character, even Article 52 list of exclusions is an imperfect basis (e.g. what are aesthetic creations and computer programs doing in the list when they can be technical?). It seems that EPC should have been revised a long time ago by clearly introducing a condition of technical character.
Concerning the definition of technicality, the Enlarged Board refers to Rote Taube case law (i.e. a judgment of the German Supreme Court), but only as a starting point. In Rote taube the invention has been defined as “instruction for planned action to achieve a calculable causal result through the application of controllable natural forces“. A requirement of mastery of the forces of nature that has been confused with the requirement of a transformation of Nature.
Then, the Enlarged board states that technicality does not imply an impact on a physical entity. Finally, the Board proposes an open conception of technicality independent of the material world. This proposal is interesting and is the first to our knowledge, although the Enlarged Board insists, wrongly, that a definition of technicality would be too restrictive and not evolutionary. This is wrong because it would not necessarily be the case and a definition would improve legal certainty, which remains a fundamental objective.
Technicality assessment
Concerning technicality assessment, the Enlarged Board applies the further technical effect method to a simulation. But above all it defines this effect, which is unprecedented. The Enlarged Board thus states: “technical effect going beyond the simulation’s straightforward or unspecified implementation on a standard computer system”.
Unfortunately, the Board does not shed any light on how the different assessment methods are linked: in particular the further technical effect with any hardware approach (i.e. “Hitachi”). It seems, however, that the further technical effect is made for exclusions claimed as such as a product, such as computer programs, as it appears for AI in EPO guidelines.
If this assumption is retained: does this mean that we could apply both further technical effect approach under article 52 and then Hitachi under article 56, when a product and a process are variations of the same invention?
In any case, eventually, the Enlarged Board adopts a vision which seems to be too imprecise to obtain the desired legal certainty in the field of computer science. Perhaps the European Union could finally take up the topic once again with AI, as it did with biotechnologies in the past? Who knows?
________________________
To make sure you do not miss out on regular updates from the Kluwer Patent Blog, please subscribe here.
It is to be hoped that this provocative piece attracts comment. I have only done one quick skim and have not yet read through G1/19 itself yet, but I offer a comment in an effort to get the thread up and running.
The author laments the absence of a definition of “technical” and asserts that this absence is to be regretted because an opportunity was lost, to increase legal certainty. I disagree. Those who wrote the EPC knew better than to include in it a definition of “invention” (or “obvious” or “inventive”) because they appreciated that no definition would be good enough to withstand the onward rush of “technology” for the next 50 or a hundred years. Further, I am sceptical that any definition would raise legal certainty to a level higher than the one already attained. My gut feeling tells me that, to the contrary, any such definition, being expressed in mere words, will lack 100% clarity and will immediately be exploited by assiduous advocates to cast doubts and erode such legal certainty as we have already attained.
There is a good reason why Art 84, EPC is not a ground of attack on validity after issue. Every claim that was ever granted by the EPC can be attacked for its level of clarity being less than 100%. Any definition of “technical” should be abjured, for the same reason.
Dear Max Drei, thanks for keeping this thread alive.
It seems the word « technical » is treated not just at the EPO, but within the entire patent sphere, much like this very important legal test, the elephant test, which goes like this : « I can’t describe an elephant, but I know one if I see one. »
It is noteworthy that the EPA in G3/08 turned down requests for a definition of « technical ». It can be said this was wise given the pace of changes in technology of this last decade. A more recent TBA decision (T 144/11 of August 2018) acknowledges that the distinction technical/non-technical is a « gray area », and recent TBA decisions struggle to provide insights to practitioners in terms of predictability as they mix technical and non-technical factors in their analysis. Constant expansion of digitisation to all areas of activity is certain to multiply the technical/non-technical intrications.
I am skeptical of efforts toward a unified positive definition of what is « technical » or of a « technical character », for the simple reason that the term « technical » encompasses different meanings, if it is based on the negative definition in Article 52(2) which is a disparate list of exclusions grounded on different policy objectives.
Exclusion of mathematical methods aims to ensure free access to the public for the common good and relates to an abstract/concrete divide.
Exclusion of business methods can be justified by the perception that there is no need to create incentives for innovations in these field.
Exclusion of aesthetic creations, software and plant varieties is justified by the existence of separate legal instruments covering these creations and the concern about cumulative protections.
As to the definition of an invention quoted from the Rote Taube decision “instruction for planned action to achieve a calculable causal result through the application of controllable natural forces“, it is not clear it can cover inventions using AI, as the outputs generated by an AI system are considered unpredictable for the creators of the system, owing to the black box character of neural networks.
.
Dear francis hagel, I enjoyed reading your answer, especially its mention of unpredictability. It used to be supposed that the whole of chem/bio was “unpredictable” but it is becoming ever less so, right. So just as what has “technical character” is evolving, so is what is “predictable”.
The patent statute in the USA confines to “the useful arts” eligibility for grant of the exclusive right of a patent. Personally, I see that as more or less a synonym for the “all fields of technology” of GATT-TRIPS. But many US patent attorneys like to argue that it is much wider than that even though, as you point out, the case for patentability of business methods has yet to be made. Patents are restraints on free trade. Any creep of patent rights into fields where they deliver no incentive to progress or prosperity should be resisted, I think.
The US Supreme Court is careful to avoid defining what is meant in the Statute by “useful arts”. Probably wise, I suppose.
Dear Max Drei,
Thanks for your kind words.
The situation in the US tends to switch from an extreme to the other, this applies in particular to patent law. The CAFC has become pro-patent at the turn of the 1980’s because of the competition of Japan, and we have seen the grant of business method patents. NPEs in the 2000’s has logically targeted big industry and banks because of the expectation of big awards but this has triggered spectacular changes in all areas of US law (eBay, KSR, Alice etc) very unfavourable to NPEs. This is just to say that the developments of patent law are deeply dependent on global factors. Compared to the US, patent law in Europe has shown remarkable stability (regrettably for some colleagues), and the EPO deserves credit for that.
Indeed. Swings from one extreme to the other. Just now, we read that Judge Wallach at the CAFC is to retire and the talk on the blogs is of whether his replacement will be a pro-patent or an anti-patent person. Hard to imagine such a debate in Europe (at least if it’s at the EPO or in England).
Perhaps it would be different if the EPO’s Boards were to decide infringement cases and if the members could write dissenting opinions, because the “global factors” you mention must inevitably influence thinking also in the patents courts of Europe.
“what are aesthetic creations and computer programs doing in the list when they can be technical?”
Juge Jacob once said that the “technical” excuse of the EPO was a restatement of the same problem. In other words, it’s a loophole to make software patentable anyway, despite the exclusion in the law.
In my view G 1/19 – fortunately – does not touch the overarching principles of technicality assessment discussed by the author. Those are well-defined and reliable since G 3/08. The further technical effect on the one hand and the any hardware approach on the other hand are complementary. They don’t need any link.
Applied to the question raised by the author in the one but last paragraph: If you need the further technical effect to pass the first hurdle, you anyway likely have no basis for Hitachi unter Article 56 (otherwise you would have used that for the first hurdle). More often – even for the process, and always for the product –, Hitachi will help you over the first hurdle and you will need some new ideas for the second.
What G 1/19 may (and was expected to) clarify was, whether the effects of a simulation can be technical – yes they can. And under what circumstances – I feel it’s slightly more generous than COMVIK, but yet have to sort it out for myself.
I am not sure what the aim of the author of the blog wanted to reach when he wrote his blog.
I fully agree with Max Drei, and to be honest I find also the blog unnecessary provocative. It is a nice play on words but does not help whatsoever.
It reminds of a joke. What is the difference between a lawyer and a scientist? For the scientist the current flows or does not flow. For a lawyer it might flow or not depending on the intended result.
I do not think that by simply coming up with a new word like “technicality” the problems with CII will be solved. A bit of common sense when it comes to understanding what “technical” means could help a lot!
What is very often forgotten is the original reason as to why computer programs where considered not patentable. It is simply because no significant search can be carried out when the disclosure resumes itself to a program listing. There are so many different programing languages with all their own syntax and rules, that it is impossible for a search authority to make prior art searches for those types of inventions. For genes a solution was found: electronic filing of sequence according to a specific format.
Only short excerpts of programs have ever been accepted at the EPO, cf. early editions of the Guidelines. Long listings of sometimes hundred pages had to be deleted before grant, exactly like the “clauses”. For all CII inventions, the inventive idea had to be clad in everyday language. Cryptography or error correcting codes are mainly mathematics, but they can have a direct technical effect, like requiring less bandwidth during transmission or storage space. That is to me perfectly patentable.
I think the case law of the BA has brought a coherent set of conditions as how to protect CII. In my view I did not really understand the necessity of the referral in G 1/19. Over all it has not brought matters fundamentally forward.
Another point which should also not be forgotten is that Art 52(2) should have been revised in what was called a second basket after the Diplomatic conference of 2000. Nothing has happened in this respect and we have seen the EBA coming up with a dynamic interpretation in order not to have the necessity of amending Art 53(b) and simply allowing the introduction of R 28(2).
Last but not least, calling for the EU to solve the problem of the CII at the EPO is somehow surprising when it comes from a lawyer working in a firm of EP representatives. It is also ignoring that the EPO has many more member states as the EU.
It is for good reasons that the founders of the EPC took their distance with the EU as the member states of the EU, or the Common Market as it was called at the time, were unable to reach an agreement upon a “European” Patent for the common market. No less than four attempts were made, and it is only the pressure of the PCT which made them move and end up with the EPC. They did not achieve this goal either with the UPC which, should it ever come into force, will not be valid for all EU member states.
“Cryptography or error correcting codes are mainly mathematics”
Or they are computer programs.
But whether it save CPU/bandwidth ressources of the computer does not matter at all. As they are just different mathematics (or a different computer program).