"I'm only against patent maximalism, seeing what it tends to lead to. I've been writing on the subject since I was about 20. It's not new to me."As we mentioned last week, nowadays there are patents on life and genetics (or genome, DNA, you name it). For life? On life? Whose monopoly? Who even 'invented' life? Religious people might say something different from and less scientific than evolutionists, but that's not the point; the question of patenting such things is one both groups can agree on. Watch what the EPO is doing:
Ares Genetics GmbH – a subsidiary of molecular diagnostic developer Curetis N.V. has reportedly been granted a patent from the European Patent Office (EPO) dubbed ‘Genetic Resistance Testing’.
"Good luck explaining to patients that patents are more important than them!"A site called "Mighty Gadget" (never seen it before!) has just published this pure spam for Withers (with a link even!), a law firm that's promoting illegal patents on algorithms. It's promoting the "hey hi" (AI) nonsense, citing "the WIPO Technology Trends report, which also reveals that IBM currently has the largest portfolio of AI patent applications with 8,290." Fake software patents in Europe disguised as "hey hi"/"HEY HI" (we need to ridicule these buzzwords) by the Campinos/Battistelli regime (the European Patent Office and US Patent and Trademark Office both use this buzzword these days, but the former seemingly influenced the latter before WIPO did the same). To quote:
Since the emergence of rudimentary artificial intelligence (AI) technology in the 1950s, nearly 340,000 related inventions have been filed for patent.
This is according to the WIPO Technology Trends report, which also reveals that IBM currently has the largest portfolio of AI patent applications with 8,290.
But just how patentable are AI inventions and machine learning, and how does the process work in Europe?
What are the Guidelines for Examination with the EPO?
As we’ve already said, the number of AI-based patent applications is continuing to rise, with the European Patent Office (EPO) also revealing an increase in the amount of applications pertaining to programmed computer inventions.
Given this slew of AI and machine learning patent applications, it should come as no surprise that the EPO has recently updated its Guidelines for Examination. Many would argue that this update was long overdue, with attorneys suggesting that some patents have been unnecessarily delayed due to outdated and analogue guidelines.
The new body of legislation makes it clear that the EPO intends to treat AI and machine technologies as forms of mathematical method. Mathematical methods currently appear on the list of non-inventions defined by art, which are technically unpatentable due to their nature.
However, a mathematical method may also be tied to the control of a specific working system or process, and in this respect it can gain technical character.
In this instance, a mathematical method moves out of the exclusion zone and into the realm of being a patentable invention, and this provides formal clarification of the position that the EPO has always adopted.
"If patent law exists to advance science, then at the moment it does a truly terrible job."The problem we have at the moment is that large patent offices, especially IP5, grant a lot of patents that courts insist should not be granted (but continue to be granted anyway). This vastly reduces legal certainty that's associated with patents, so now the patent maximalists try to manipulate politicians, wrestling the laws and bribing people like Coons. In Europe we can see that in UPC. "Explanation given" wrote the following comment last night on the breaking news of a very high British court squashing a key European Patent, noting that patent troll "Conversant's case is dead if the patent is invalid for added matter, but he still decided on obviousness and sufficiency (and infringement) so that the case can be dealt with in full on appeal."
Here's the full comment:
There is nothing in the EPC that prevents a claim from being new and inventive while at the same time infringing Art. 123(2). Just apply the law as it is written. The "effective date" for determining the state of the art for the claimed subject-matter is determined by Articles 54 and 89 EPC. There is nothing in those provisions that prevents the EPO from establishing the relevant state of the art if a claim happens to infringe Article 123(2).
"The part of the claim relating to added subject-matter should however not be there and no effective date can be attributed to it."
This is pure myth that unfortunately has taken root in the minds of a few EPO examiners. Please cite one passage in the Guidelines or in the Case Law book that supports your position. Again, it is Articles 54 and 89 that determine the relevant state of the art, and those provisions do not require the claim to comply with Article 123(2).
"At the EPO, a claim comprising added subject-matter leads to a refusal in examination or to a revocation in opposition, cf. G 1/93. Such a claim is never compared to any prior art, as it does not have a raison d'être."
This arrogantly assumes that an opposition division's decision on 123(2) is always correct and will survive appeal, which has no basis in reality.
If the patent proprietor has an at least arguable case on 123(2) (that did not convince the OD but still might convince the BoA), there is nothing to stop the OD from continuing with inventive step (even though it is not obliged to do so). Then if the BoA on appeal agrees with the patent proprietor on 123(2), a remittal for examining inventive step has been avoided.
This is how the English courts deal with cases. Justice Arnold was well aware that Conversant's case is dead if the patent is invalid for added matter, but he still decided on obviousness and sufficiency (and infringement) so that the case can be dealt with in full on appeal. There is no legal reason why EPO divisions could not do the same.