It has already happened in the US and it may soon happen in Europe

Summary: Looks like EPO judges might soon be able to rule against software patents -- a pressing matter and another fresh opportunity to set things straight in Europe (like Alice in the US half a decade ago)

THE European Patent Office (EPO) cannot help itself. Yesterday it wrote: "Tomorrow is your last chance to register for this event in Lundun, Sweden where we will address the topic of computer-implemented inventions in #Medtech with a focus on the value of patents for SMEs: https://bit.ly/2E5l03D #startups pic.twitter.com/sG6xVSbnhG" (accompanied by the same stock photography it used almost half a dozen times lately).

Notice how they add words/buzzwords like "med" and SMEs"; now there's even "startups" as a hashtag. We remarked on these earlier this week and last week. By "computer-implemented inventions" the EPO means patents that ought not be granted and this tweet was apparently deleted later on the same day (we cannot find it anymore). The EPO was also promoting driving-centric software patents in Europe. Yesterday it wrote: "Europe and the USA have a strong lead in self-driving vehicle innovation with about 1 400 European patent applications each in 2017 alone. See how other regions performed here: http://bit.ly/SDVstudy #SelfDriving #FutureOfCars pic.twitter.com/L0argQl72d"

It's that "SDV" nonsense that we responded to like a dozen times before. Having come from that area myself, as a developer, it's clear to me that the vast majority of these would be algorithms. "SDV" is just like "CII" or "4IR" -- it's a buzz phrase in acronym form. It's like "AI" -- a term which according to new research is being grossly misused (see our daily links). Many things that companies describe as "AI" are just ordinary algorithms and many so-called 'AI' startups have nothing to do with "AI" (hence the scare quotes).

We have come to the point where the EPO generally compensates for granting illegal or bogus patents by sticking misleading labels on these. The U.S. Patent and Trademark Office (USPTO) has copied some of these methods, e.g. calling software patents "AI". Things have become rather different in the United States, where software patents are otherwise very difficult to assert in court. The patent maximalists have totally, entirely lost the plot. They're panicking. They feel greatly irritated. Consider this new tweet from a retired attorney. That ridiculous patent maximalist, Janal Kalis, generalises; he now calls "anti-patent" anything that merely opposes abstract patents or patents on algorithms (which no software developers ever wanted) and also attributes anything against software patents to "EFF". His tweet says: "The USPTO is seeking comments on the 2019 Revised Subject Matter Eligibility Guidelines. So far, 103/119 (87%) of the comments come from the EFF and are anti-patent. If you have another position, please send your comments to the USPTO asap: Eligibility2019@uspto.gov"

They panic. The public (not a few law firm) is talking and is now controlling the debate. Iancu (the USPTO equivalent of Ajit Pai at the FCC) cannot hide the fact that citizens of the United States oppose his plan.

How long (or how much longer) can judges be ignored, snubbed and ridiculed?

Gene Quinn of Watchtroll has returned for a day and earlier this week he pretended that the Federal Circuit (CAFC) has not adopted the SCOTUS decision on TC Heartland even though it clearly has. We don't suppose fact-checking even matters in that horrible propaganda site, which smeared judges for merely following 35 U.S.C. ۤ 101. This is what it nowadays boils down to. It certainly seems like the only CAFC reversal Watchtroll is able to find this week (in favour of plaintiff) has nothing to do with patents; it is about trade secrets.

This is of course encouraging. It doesn't seem as though a comeback for software patents in the US is possible (now or ever). The judges make the final call, not the attorney whom Trump put in charge of the Office (or the French/Portuguese banker whom Battistelli chose for Europe).

Here in Europe there's an upcoming case/referral, which we wrote about thrice before [1, 2, 3] and last mentioned on Wednesday. It can, in theory or in principle, become a 'European Alice' though it requires judicial independence (which is lacking).

IAM (blog) has just published, without the usual paywall that protects patent trolls from scrutiny, an outline from Philip Naylor regarding software patents. As usual, the patent trolls' lobby pushes for (in favour of) software patents in "Patent cases in the EPO and UKIPO: different qualifications for computer software" (algorithms).

Much of the status quo gets attributed to an IBM case from two decades ago:

As a result, patents can be granted at the UK Intellectual Property Office (UKIPO) and at the EPO for inventions which involve the categories of excluded subject matter set out in Article 52. However, the “as such” qualification has also been used to limit the scope of allowable subject matter in these categories.

Over the years, the EPO and the UKIPO have developed the way in which the allowability of patent applications for computer software and business methods is assessed. The EPO has settled on an approach that gives greater certainty to applicants in comparison to other patent offices, such as the USPTO, where the law in this area is in a state of flux. The UKIPO’s approach is similar to that of the EPO, but with some subtle differences.

[...]

In 1999, two EPO Technical Board of Appeal cases, T935/97 and T1173/97 (IBM), established that claims having the form “a computer program product” and “a computer readable medium having a program recorded thereon” were allowable.

Following the IBM decisions, it became accepted practice at the EPO that software could be claimed directly. This continues to be its practice so long as other requirements (ie, clarity of the claim language and sufficiency of disclosure) are met.

In Case T931/95 (Pension Benefits System Partnership), it was held that the claims of an application must define non-excluded subject matter and be novel and inventive. The claims were directed to a method and an apparatus for controlling a pension benefit programme. The method claim referred to technical means, but was refused because it related to a method of doing business as such (ie, it did not define non-excluded subject matter).

In this case, the apparatus claims did define technical features and were not refused merely because they related to excluded subject matter. However, the apparatus claims were refused on the grounds that they lacked an inventive step because it was viewed that the differences from the prior art lay in an economic field (ie, non-technical) and hence there was no technical contribution provided by the distinguishing features of the invention. This case marked a substantial shift in the EPO’s approach to non-excluded subject matter.

[...]

In 2008 in referral G3/08, the president of the EPO referred questions regarding the patentability of computer programs to the EPO’s Enlarged Board of Appeal. After a lengthy period in which many amicus curiae briefs were filed, the board declined to answer the questions posed in the referral on a legal technicality; it was held that, since there was no divergence in existing case law, the legal requirements for the referral itself were not met.

Since then, the EPO has continued to apply the precedents set out in the Pension Benefits, Hitachi and Comvik cases, among others.

There may soon be another new precedent in the form of "simulation" software. We hope that the judges can regain their independence some time soon and rule on the matter without fear of retribution. ⬆