Bonum Certa Men Certa

The War on Patent Quality

Poppies



Summary: A look at the EPO's reluctance to admit errors and resistance to the EPC, which is its very founding document

THE companies that make weapons want war. Firms that provide representation in lawsuits want a lot of lawsuits. That's just common sense. But the profit motive is orthogonal to morality. Patent law should adhere to morality or to science, not the profit motive. Can it withstand lobbying and subversion?



"Firms that provide representation in lawsuits want a lot of lawsuits."Gün + Partners, who habitually write at the site of the patent trolls' lobby (IAM), maintain the illusions associated with patents because over in Turkey they profit from patent applications (even if futile, worthless). "First published by IAM – Patents in Europe, 11.04.2019," as it says at the bottom, days ago the following got (re)published in Mondaq, mentioning the EPC and the EPO as follows:

Yes, there are specialised civil and criminal IP courts in Turkey. The specialised IP courts are located in Ankara, Istanbul and Izmir. For other cities, the first-instance civil court deals with IP-related disputes. The judge of the first-instance civil court has a basic knowledge of intellectual property; however, most cases are referred to a court-appointed expert panel due to the expertise and technical knowledge required for the dispute.

[...]

As Turkey is a party to the European Patent Convention (EPC), and the decisions of the EPO on European patents are valid and binding for a Turkish validation of that European patent under the European Patent Convention 1973, the decisions rendered by the EPO Opposition Division or its board of appeal are usually submitted to the court by the parties as persuasive evidence if the Turkish validation of an EPC patent is the subject of the dispute.


The problem at the moment is, the EPO departed from the EPC (in many respects). The EPO is now a lawless, out-of-control institution. Nobody likes this institution except top-level management, which is robbing it in the absence of oversight.

Even managers at the European Patent Office are utterly sick of the situation (even nearly a year after António Campinos took over). Just see the image in this new blog post likely composed by an insider; it's mostly red, which is rather revealing, accompanied/conjoined with the following succinct text:

A single picture from the survey summarizes the present situation at the EPO. In red, people are unhappy about the situation, in green people find the situation to their advantage. G13 is the highest grade an examiner can have, the people in G14, G15 and G16 are managers. They are also the people whose remuneration was considerably increased under Battistelli leadership.


Line managers and Directors are said to have attended EPO protests. Battistelli managed to make it a tyranny of very few; only Team Battistelli and patent trolls' representatives are happy. The Team UPC types...

"The EPO is very much ‘in bed’ with patent trolls, not just with European software patents that these trolls leverage."One German patent attorney has just quoted [1, 2] or paraphrased a judge: "At a seminar at the Munich-based Max-Planck Institute for Innovation and Competition judge Fabian Hoffmann of the patent senate of the @BGH_Bund (FCJ) considers not granting injunctions against patent infringers if the patentee is not a competitor, i.e. NPE [troll ...] He concludes the presentation with his belief that patent law - the purpose of which is to protect innovation - should be open to innovation itself."

Not a bad idea actually. However, over the past week the EPO promoted LESI almost every day (in Twitter). The EPO is very much 'in bed' with patent trolls, not just with European software patents that these trolls leverage.

That same German patent attorney has just noted that: "The EPO will stay all pending examination and opposition cases in which the decision depends entirely on the outcome of the referral G1/19 ("Patentability of computer-implemented simulations"). https://www.epo.org/law-practice/legal-texts/official-journal/2019/04/a35.html … https://patentlygerman.com/2019/03/01/epo-enlarged-board-of-appeal-to-decide-on-the-patentability-of-computer-simulations/ …"

Other patent maximalists in Germany also note that the EPO now acts as though software patents might soon be flushed down the toilet (which we strongly doubt, but one can always hope for 'European Alice'). Here's this new article:

EPO Stays Examination Of Applications Relating To Technical Simulations In Light Of Enlarged Board Referral



As reported recently, an EPO Board of Appeal has referred questions to the Enlarged Board of Appeal seeking clarification on the extent to which a computer-implemented simulation of a technical system or process can solve a technical problem by producing a technical effect and thus contribute to an inventive step. The referral is pending as case G1/19.

In the underlying case (T 0489/14), the Board was minded to disagree with the existing case law in this area and so the Enlarged Board's answers to the questions referred could potentially overturn this well-established case law.

The EPO has now issued a notice in which it immediately stays all cases currently in examination or opposition whose outcome "depends entirely on the outcome of the Enlarged Board of Appeal's decision". This is a standard practice that allows time for the law to be reviewed without risking depriving applicants of their rights either via refusal of their application or the grant of an invalid patent. However, as Enlarged Board referrals typically require one or two years to resolve, it also introduces a significant delay.

The notice makes it clear that the stay will only be applied where the assessment of inventive step requires deciding whether or not a computer-implemented simulation can be considered to produce a technical effect which goes beyond the simulation's implementation on a computer. If cases can be examined without consideration of this issue (for example because there are other inventive technical features in the claims), then the stay will not apply. The notice also indicates that, solely for the purposes of considering whether or not to stay proceedings, "the term 'simulation' is to be interpreted as meaning an approximate imitation of the operation of a system based on a model of that system", based on point 21 of the Reasons in the referring decision.


The EPO actually needn't rely on a referral to know that software patents aren't compatible with the EPC (and never mind the judges' lack of independence herein). Yesterday the EPO once again spoke of "AI-based inventions"; the EPO just means software patents which are illegal/invalid. To quote the tweet in question: "Our next study visit will introduce attendees to the complex field of AI-based inventions and the new problems they pose for our patent examiners."

And on Friday morning (same day) World Intellectual Property Review published this typical EPO puff piece. Where's the scrutiny over patent quality? These publishers have so much interest in puff pieces, not scandals or newsworthy stuff, instead just PR.

Another German patent attorney, Dr. Thorsten Bausch, has characteristically (for him at least) blasted the EPO for chaos. He split his post into several different sections.

Here's the part about quality being conflated with speed (a subject we covered here dozens of times in the past):

Evidently, the fundamental wisdom provided by Erich Kästner also applies to the subject of quality. While the new President displays much more openness to input from stakeholders on this subject than his predecessor, actions would speak louder than words. On the whole, I wish I had something more substantial and tangible to report here.

I and many others have extensively argued that quality of EPO products (e.g. searches and decisions to grant or refuse an application) must and will suffer if unrealistic production targets are set and are made the key parameter against which staff performance is measured. The VPP (the German Association of Intellectual Property Experts) discussed quality in its recent Spring Meeting in Bamberg and invited me to elaborate a bit further on this topic. My slides are available here (in German). They may give readers some insights into where I see a particular risk of quality erosion and which counter-measures I would suggest the EPO take. As always, I will welcome any comments and critique.

There was a lively discussion at the end of my lecture, during which a former EPO examiner stood up and explained to the audience how lengthy and cumbersome the process towards a decision of refusal is, compared to allowing the application to proceed to grant. Not only that, but you also make yourself unpopular by drawing two more of your colleagues into this struggle. I agree, but would add that a lack of adequate time may occasionally also lead to errors or misjudgments to the detriment of the applicant, e.g. if Art 123(2) is applied too photographically, or if the examiner introduces errors in the Druckexemplar at the Rule 71(3) stage.

Mr. Niclas Morey, the EPO’s Principal Director of Quality, held the next lecture after mine. Contrary to my view, he was very adamant that speed is part of quality. While I appreciate that stakeholders expect the EPO to act timely and that even a very good official action is not particularly useful if the first one only comes after 10 or more years, I do not see the point in mixing the categories of quality (in content) and speed. For sake of intellectual precision and for very practical reasons, I think that a narrower understanding of quality would be more appropriate. Quality essentially describes the properties of a product and measures it against expectations in categories of good-bad, rather than in quantitative categories (fast-slow or high-low speed or turnover). A distinction between the two also helps to better understand and accept the inevitable trade-off between quality and speed: the more you produce within a certain period of time (with the same resources, as the EPO has done for the last 5 years), the more errors will occur and the more quality will suffer, unless something specific or magic is done to lower the error rate. It is true, though, that (good) quality is only one element of customer satisfaction, the others being timeliness, adequate pricing, and achievement of the desired result (grant vs. rejection).


Notice the comment from MaxDrei, a patent attorney who used to comment in IP Kat and is nowadays commenting a little more in Kluwer Patent Blog:

As to “quality”, I will assess my new car for its “quality” in terms of its fitness for purpose, in use. How long it took to advance down the production line in its assembly plant is not within my definition of “quality” of the “product”.

As an investor in society, I would see speed of the production line as something relevant to the “quality” of the management of the production plant.

As a patent attorney, I think that the pendency of the patent application has nothing at all to do with the “quality” of the respective issued patent.


"Experienced Examiner" also weighed in on quality:

QUALITY One of our last official meetings touched on “quality”. In line with what Thorsten reports, the official stance really is that “quality” as defined by EPO management has three pillars: EPC compliance, timeliness and user satisfaction. It seems that each pillar has equal weight. In concrete terms, EPC compliance can at most define 33% of “quality”.

The numbers of internal audit – compliance rate around 75%, meaning 1 out of 4 patents should not have been granted the way it was – are therefore not an issue, as long as timeliness and user satisfaction are high enough. Our director, consequently, did not speak about “quality” but about “perception of quality”. Any resemblence to “perception of independence” is of course just coincidence.

Concerning the numbers of patents/searches, be informed that the average examiner has additional tasks. One of those tasks is classification of published patent documents. In spite of all efforts, even the newest search engines provide decent (not very good) results only when relying on intellectual, human classification. Machine classification and machine search without human input are definitely not good enough. They will likely not be in the foreseeable future, because technology develops too fast to provide enough data of similar nature for training of the algorithms. This is a huge difference to medical diagnosis: the human body does not change that fast.

Classification does not generate patents or income. It has been neglected under Batistelli, with a corresponding impact on search quality. Also, the classification schemes (like CPC or IPC) need to be updated over time. This has also been neglected. Briefly, the number of not classified documents is high and the schemes are outdated. There is now massive pressure to reduce the classification backlog, while the number of patents to be granted has dropped only slightly. In concrete terms, the average examiner does not have more time than the last years for dealing with files.

How much time the EPO foresees is laid down in “corridors”, depending on grade (aka salary) and technical area. These “corridors” are used for target setting, reporting, step advancement and I do not know what else. They have never been discussed with staff and never been officially published. I believe the patent community should know how much time the EPO foresees. Thorsten, that could be a point for you: ask VP1 to put those corridors (for search, examination and opposition) on the EPO website. The Boards of Appeal should do the same.

SOCIAL DIALOGUE The staff survey results were a disaster for higher management, starting on director level. The concrete actions I hear and see: the team managers (one level below director) are now tasked with getting their teams together and identify areas of improvement. We have been assured that “work goes on on all levels”. Whatever that work might be (I personally doubt there is any work going on), it does definitely not include to meet with staff on the working level to get first hand impressions and provide direct replies.

All things considered, I have the impression that the staff survey was done to have some glossy paper. It was and it is not intended to change anything. I’d love to be proven wrong here.

FINANCIAL SUSTAINABILITY The latest financial study is out. In a nutshell, the consultants managed to show that the EPO will be broke once again in 2038. Looks like all Battistelli measures – sacking staff representation, breaking the law, interfering with the Boards of Appeal, drop in quality (EPC compliance) – have been useless. Take a look at the details, if you get your hands on the study.

The consultants, maybe based on EPO suggestions, assumed an increase of expenses, mostly salary, a bit above inflation every year, while the fees shall be increased only in 2020 and then remain stable. Expenses increase exponentially, while income is stagnant. Clearly, with these assumptions, everybody will be broke in 20 years. I wonder why the EPO bothered doing the study in the first place.

The EPO will not disappear because someone did a shady financial study. But it may help to understand why staff has no respect and trust at all in management. It also shows the responsibility of the Administrative Council. They simply cannot trust anything even the new President, Campinos, puts on the table.


Someone by the name "Martin" (maybe a pseudonym) wrote:

The EPO may have a lower production target as a whole, but also a reduction in headcount of Examiners is foreseen (due to workforce moving into pension age).

Therefore the average number of products expected from individual examiners has risen again, and is expected by management to continue to rise according to similar levels as realized the last few years (see financial study).

Which means for many Examiners who did achieve a higher production, the evaluation in the “reporting excercise” nevertheless degraded, as they did not achieve a rise in the production as expected.


Here's what Bausch said about SUEPO:

Whatever the merits of SUEPO’s position may be, this letter somehow prompted me to check Wikipedia for the keyword „amnesty“. This proved to be a very interesting and educational reading, providing a lot of food for thought. For example, I learned (from the German version) that there was a time when it was customary for incoming French presidents to proclaim an amnesty for certain kinds of prisoners, mostly those who had been found guilty of lesser offenses, or political prisoners. The same tradition existed in England upon coronation festivities and other solemn occasions. I also learned that an amnesty may generally be extended when the authority decides that bringing citizens into compliance with a law or policy is more important than punishing them for past offenses. The general purpose of amnesty is to help end a conflict.

I do not really know what exactly the SUEPO leaders are accused of and whether they did anything wrong. Considering the EPO’s history during the Battistelli era, I have my doubts (based on verifiable facts, as the Corcoran story shows) that the severe sanctions against the SUEPO leaders were justified. But even if they were, would it not now be an excellent time for the new president to make peace with the staff union and settle the sanctions against its leaders? At least I would interpret this as a sign of prudence. It would help to foster the important objective of regaining the trust of the staff, enhance their motivation and ultimately promote quality. I would argue that quality can only be sustainably kept or increased, if those who are to deliver it feel motivated, rather than threatened and placed under pressure.

Mr Campinos has now been in power for almost one year. He has proclaimed that he is interested in restoring the social dialogue in the EPO. Now would be a good time to put actions to words.


Laurent Prunier shows up in the third comment (scroll down a little), noting that the punishment imposed on him by EPO management was based on a fabricated allegation or 'violation' -- brought forth from Team Battistelli, the thugs who now head (or hide in) CEIPI.

Here is Prunier's official response (reproduced again):

As the person directly concerned, I am responding to Mr Battistelli's letter to you and would appreciate if you could publish this answer so that your readers can be fully informed.

I deny having ever harassed or defamed anyone (nor have I seen any of my fellow colleagues, staff reps and/or SUEPO officials harassing or defaming anyone).

The alleged "victim" did NOT file a complaint against me. The person who filed it was a very close associate of Mr. Battistelli.

The staff representatives in the disciplinary committee have not found that I was guilty of harassment. That finding was 3:2, only by management side.

The easiest solution for the public to assess the truth vs. story-telling is for Mr Battistelli to lift the confidentiality he imposes on me and I will gladly publish all the documents.

Bien cordialement - Best regards

Laurent Prunier



Thankfully, Dr. Bausch has also taken note of the 'heist' I told him about the other day, as per an anonymous blog pointing to the EPO's own documents -- documents that it apparently censored quickly afterwards. Here's a comment about this 'heist':

EPOTIF is not related to pensions. Pensions are covered by the RFPSS – which too saw short term losses. Short term losses are normal for any pension fund.

What was not normal in the past, was the way such losses used to be hidden in EPO accounts by using a “corridor” approach which only required reporting gains and losses if they fell outside a range of outcomes. The yearly gains and losses are now visible, which is good for transparency, but apparently bad for the nerves.


How much money is the EPO willing to lose for likely Battistelli gains? A billion euros? When will anyone -- let alone Campinos -- hold Battistelli accountable? He destroyed an excellent patent office so quickly and he personally gained enormously from that.

Recent Techrights' Posts

X.Org is Still Not Dead
Oracle still developing it
Microsoft is getting ready to cause many employees to resign
Having already laid off many workers earlier this month, it now tries another approach
"Maybe the Problem is You"
they probably felt like they had no choice because they really needed this Microsoft money
GNU OS, Powered by Hurd
Choice is good, as long as choices exist that respect the users' freedom
European Patent Office (EPO) Reformation Project
It's a stain on the EU's reputation
Slopwatch: Google News and Other Slopfarms
Google News is rewarding sites that misuse LLMs and cheat the Web
 
Links 15/08/2025: German Government Falls Short on Free Software, Russians Breach EU Systems
Links for the day
Microsoft is Still Losing Cyprus
The market share goes down, so share prices go up
Microsoft Accenture is in Trouble
For one thing, its debt doubled in a matter of months
News Will Slow Down and Slop Will Contribute to the Slowdown
In recent years every time there was some holiday or major break the number people who "came back" shrank
Upgrading IRC Network of Techrights
a new version of the daemon we've used since 2021 was released very recently
"Register Debate Series" About Microsoft in the UK is Controlled by Microsoft (US)
The Register is run by Microsoft "Analysts", so the debate is doomed from the get-go
IBM is a Terrible Model for Red Hat
"Most likely caused by laying off too many people"
Microsoft Problems in Palestinian Territory and Israel
Microsoft stock (share price) goes up when market share goes down
Slave is Not a Bad Word, We Need to Use It Sometimes
Who does such exclusion of words benefit? What sort of expression will be deemed impermissible and subjected to CoC enforcement?
National Day of Action
"This Friday, August 15th, there is an organized, petition-based, protest of Wells Fargo in major cities across the US," Richard Stallman wrote
Our Gemini Editions Now Contain 100,000+ GemText Pages
Our Gemini Editions aren't small, even if Gemini Protocol is still the 'underdog'
The Relations Between the United States and Europe Deteriorate, Should Europe Continue to Rely on American Tech Giants?
The shallow notion that made-in-USA software is fairly safe for Europe to rely to is coming to a standstill
Techrights and Tux Machines Running as Usual During Vacations
No interruptions, maybe temporarily slowdowns
Gemini Links 15/08/2025: ADHD and "Random Weird Things"
Links for the day
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Thursday, August 14, 2025
IRC logs for Thursday, August 14, 2025
"Article 52. PATENTABLE INVENTIONS" in the European Patent Convention
Some time tomorrow we'll have a complete local copy of the EPC
Serial Slopper (SS) Still at It, Still Misusing Plagiarism Tools and Cheatware for Images and Text About "Linux"
All the slopfarms are a very big problem
Reddit Deletes Stuff, But Not for Being False or Misleading
Yet another one of those articles that speak of a man in his 50s as if he's terminally ill
Times of India and India.com Are Clickbait and LLM Slop
Google continues to reward bad actors
The More "Market Share" Microsoft Loses, The Higher the Shares Go
People joke about the same sort of thing in relation to IBM
To OIN, Software Patents Are Not a Problem
Had software patents ceased to exist, OIN too would cease to exist and its staff would be unemployed.
Microsoft's Bankruptcy in Russia is Only the Beginning
Due to politics it mostly makes sense that Windows is being phased out, also in part due to policy changes
Microsoft-Funded Publishers Lied to Us About Vista 10 and Now Advocate Us Owning Nothing
They want you to own nothing, but they also want you to buy a PC on which to become Microsoft's slave and they make it harder if not practically impossible to remove Windows
Articles Promoting and Celebrating Wayland Are LLM Slop
New example (100% slop)
The Register MS, Dominated by American Editors, Says UK Should be Run (Digitally) by Microsoft US
The Register MS is sponsored by American money, run by Americans, and its chief editor is a Microsofter from the US
Gemini Links 14/08/2025: Drought, Climate Experiments, and LLM Slop Considered Detrimental
Links for the day
Links 14/08/2025: Second-hand ThinkPad and Enhanced Surveillance on Chipsets from the United States
Links for the day
Moral Standards From the Masters of Linux
They get hung up on minor language issue and promote this crazy theory that racism will go away if only everyone spoke a little differently (no matter where he or she came from)
Links 14/08/2025: Data Brokers Hiding Opt-Out Pages From Google, "Fight Chat Control"
Links for the day
FSF Infrastructure Under Constant Attack
The disconnect (literally) has had an effect on credibility
Feels Like The Register MS is Trying to Diversify a Bit
If The Register MS goes back to being The Register US (or UK), that will be a nice improvement
Gemini Links 14/08/2025: Reading Journal and LLM Fatigue Revisited
Links for the day
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Wednesday, August 13, 2025
IRC logs for Wednesday, August 13, 2025
Hopping From One Set of Buzzwords to the Next
Rotating hype and vapourware
Currys PCWorld Hates GNU/Linux Even Though It Runs the World
If more and more people choose to remove Windows, then Currys PCWorld will feel the financial impact of its dumb policies
Internet Relay Chat and Gemini Protocol Help Us Relive the Net of the Dial-Up Era
The kids were alright
The Register MS Takes More Money to Boost Slop Hype, This Time From Snyk, a Notorious FUD Source
At some stage or at some point they might even decide to stop doing so
"GPT-5" is Another Microsoft Dead Cat Trying to Bounce
The hype, the momentum (or the inertia) is wearing off
Microsoft Windows Losing Its Grip Near Turkey and Russia
The 'corridor' nations connecting Iran to Europe
Slopwatch: LinuxSecurity, Google News, and Serial Slopper (SS)
The slop, the bad, and the ugly
Links 13/08/2025: The “Incriminating Video” Scam and Corruption in South Korea
Links for the day
Gemini Links 13/08/2025: Movie Memories and Mystery Machine Bus
Links for the day
"AI" Hype or LLM Slop is Not About Efficiency, It's About Lowering Standards
It does not seem like IBM is genuinely committed to the same goals (or commitments) as the original Red Hat
Links 13/08/2025: GitHub Trouble and Openwashing by Microsoft OSI With the Typical Buzzwords
Links for the day
If Free/Libre Software is Adding Trillions in Value to the European Economy, Then the European Commission Must Crush Software Patents
Further to what we wrote yesterday
Microsoft Swallows GitHub Losses
Only Microsoft knows how much money it has already lost on GitHub
Gemini Links 13/08/2025: Climate, Coffee, and Deploying Troops in Washington DC After Pardoning 1,000+ Insurrectionists in Washington DC
Links for the day
The Register MS Lowered MS Focus This Week
We hope The Register recognises its errors and tries to make up for them
Learning Ethics From Jeffrey Epstein's Enabler/Client/Ally, Coca-Cola, and Microsoft Accenture
Whatever merits vocabulary changes initially had are being tainted or obscured by later iterations, which tell us to avoid word like "normal", which apparently offend some people (so they argue)
Personal Attacks From Rust People Serve to Confirm They Have Lost the Argument
"The discussion I find around the net so far has no technical merit and centers around ad hominem"
Physical Meters and Purely Mechanical Meters Aren't Dumb; It's Dumb to Mock or Dismiss Them as Antiquated
I've learned a lot this week, both online and over the telephone
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Tuesday, August 12, 2025
IRC logs for Tuesday, August 12, 2025