CASELAW-EPO - reviews of EPO Boards of Appeal decisions

G 2/21 – Some thoughts on plausibility in view of the communication of the EBA

chat_bubble 8 comments access_time 6 minutes

In preparation of the OP to be held on 24.11.2022 in case G 2/21, the EBA has published a communication pursuant Art 13 and 14(2) RPEBA.


The question of plausibility is very important in chemistry/pharmacy and it is thus comprehensible that those industries want to keep the free evaluation of evidence as much as possible.

Especially in second medical use claims the effect is part of the claimed invention and is thus of paramount importance when assessing inventive step but also when assessing sufficiency. As the effect a functional part of the claim, then sufficiency is at stake and not inventive step.

The amicus curiae filed

When looking at the 20 amicus curiae some tendencies are discernible.

Whatever the replies say otherwise, the free evaluation of evidence should be kept intact and by no means be curtailed by the decision of the EBA in matters of plausibility.

One writer is however of the opinion that legal certainty should exclude post-published evidence.

On the one hand speculative patenting should be avoided, especially in view of the first to file system applicable at the EPO.

On the other hand it is the essence of the problem-solution-approach (PSA) which can be based on prior art not known to the applicant/proprietor at filing which leads to the necessity of filing post-published evidence in order to assess IS. The formulation of the objective technical problem could lead to express different effect(s) to that(those) mentioned in the original application.

All technical domains should be on a par and no specific rulings should be established for specific technical areas, especially for the pharmaceutical industry.

One amicus curiae summarised the situation quite nicely: the approach to plausibility should lie between the two extremes of excluding post filed data completely and not taking plausibility into account at all.

In general, the answer to question 1 should be No which renders any answer to the questions 2 and 3 superfluous.

Some also proposed to rewrite the questions.

I have some sympathy with this position as the question referred only to plausibility with respect of inventive step and did not broach the topic sufficiency.

Next to the 20 amicus curiae there was one observation of a third party. Like in a plurality of the former, the third party mentioned the effect resulting from the application of the PSA to be taken into account. The third party considers that mere allegations in the description cannot be taken into account to identify an effect; corroborating evidence is required.

Position of the President of the EPO

The president of the EPO considers the referral admissible but is of the opinion that answers are not only required under Art 56 but also under Art 83.

Ab initio plausibility is a question to be addressed by the applicant/proprietor and ab initio implausibility is a question to be addressed by the opponent.

New effects might be relied upon provided that the effects are implied or related to the problem initially suggested.

The EPO must be put in a position to establish whether there is any technical effect at all, i.e. whether the technical effect is achieved and whether the subject-matter was non-obvious compared to the state of the art.

No exception to the principle of free evidence is actually needed.

Depending on the prior art, the re-formulation of the technical problem could lead to the objective technical problem being less ambitious than originally envisaged by the applicant.

The EBA’s communication

For the EBA, the referral appears admissible.

For the EBA, the principle of free evidence is a fundamental principle which cannot be ignored and is recognised in the contracting states and is enshrined in Art 113(1) and 117(1).

The EBA intends to provide guidance on the issues referred to in questions 2 and 3.

The EBA does however not intend to go further than the referral and only intends to reply in the framework of the assessment of inventive step.

Neither does the EBA see the need for the referred questions to be re-drafted or re-ordered.

The Enlarged Board recalls that in the light of the case law of the EPO and the Contracting States, what matters is what technical knowledge the person skilled in the art, with his general knowledge, would understand at the filing date from reading the application as filed. The technical effect relied upon, even at a later stage, must be encompassed by that instruction and embody the same invention.

For the Enlarged Board, the question is therefore whether, in view of this technical teaching of the application and the general knowledge, the person skilled in the art would have significant reason to doubt the alleged and invoked technical effect.

In the absence of such doubts, the decision-making body should be able to rely on post-published evidence to decide whether or not it is convinced of the presence of the technical effect.

If, on the other hand, the person skilled in the art has significant doubts, the possibility of using post-published evidence successfully seems questionable.

It appears the the EBA will, like for instance in the disclaimer decision G 1/03, endorse the existing practice of the EPO and its boards of appeal.

Comments – Inventive step and sufficiency

The EBA made clear that it only intends to reply to the questions raised in the framework of the assessment of inventive step.

On the other hand, even if the referral only raises the question of plausibility when assessing inventive step, the same criterion also applies when assessing sufficiency. This is also the position of the opponent in the referring decision T 116/18.

It is thus difficult to follow the EBA on this point.

In my opinion the questions of inventive step and sufficiency are linked, especially in view of G 1/03 and T 2001/12. This is also the position of the opponent in the referring decision T 116/18

In G 1/03 the Enlarged Board of Appeal stated the following (See Reasons, point 2.5.2, last §):

“a lack of reproducibility of the claimed invention … may become relevant under the requirements of inventive step or sufficiency of disclosure. If an effect is expressed in a claim, there is lack of sufficient disclosure. Otherwise, i.e. if the effect is not expressed in a claim but is part of the problem to be solved, there is a problem of inventive step

T 2001/12 is not an isolated decision and has exemplified the interplay between Art 83, 84 and 56.

What is at stake in T 2001/12 is the effect claimed by the applicant in its original disclosure. If said effect is not credibly (=not plausible) achieved with the information originally disclosed but not claimed then the objection should be lack of IS. If the effect is not credibly (=not plausible) achieved but claimed, the objection should be lack of sufficiency.

In the referring decision T 116/18, the effect is not claimed as such and hence according to G 1/03 the question at stake is thus a question of IS and not of sufficiency. The effect is not the effect resulting of the problem-solution approach, but the effect as originally disclosed in the application as filed, cf.  § [007] in EP 2 484 209 A1.

I am thus of the opinion that, in spite of the referral only aiming at the assessment of IS, answers should also be given for the assessment of sufficiency.

It might be true that in application of the problem-solution-approach (PSA) the applicant/proprietor might be faced with new prior art which was unknown to him.

It remains nevertheless that any invention is a technical solution to a technical problem. The question at stake can be formulated as follows: which is the technical problem to be taken into account when deciding on IS or sufficiency. Is it the original problem disclosed in the original application (or which can be interred from the latter), or the objective technical problem resulting from the PSA.

Reliance should be given to the problem which was initially disclosed in the application (or which can be interred from the latter), and not the objective technical problem resulting from the application of the PSA.

Although it was for assessing added subject-matter when criticising the “essentiality test” (now the three point test) in T 2311/10, the board made clear the “problem” to be taken into account in step 2) of the then “essentiality test” is the problem as defined originally in the application and not the “objective problem” when it comes to asses inventive step.

The situation is similar when it comes to IS or sufficiency. What matters is the originally defined problem and not the objective problem resulting from the PSA.

The technical effect relied upon, even at a later stage, must be encompassed by this problem and effects originally stated and embody the same invention. This is a difficulty which it will be interesting to see whether the EBA will discuss or will ignore.

It is clear that depending on the prior art, the re-formulation of the objective problem could lead to the objective technical problem being less ambitious than originally envisaged by the applicant.

Even less ambitious, it should not allow to later claim effects or advantages which were not originally disclosed.

This is also the reason why the acknowledgement of prior art to be introduced in the description after examination should remain purely factual and not relate to effects or advantages not originally disclosed. Anything else would end up with an objection of added matter. Here we would be back to T 2311/10.

It is to be seen whether the EBA will maintain its strict stance and limit its replies to the assessment of IS, including the effects resulting from the objective technical problem, or if he will also deal with sufficiency.

The OP before the EBA on 24.11.2022 promise to be highly interesting.

It is foreseen that the OP will be in-person on the premises of the BA in Haar (Room CONF201-auditorium).

Share this post


8 replies on “G 2/21 – Some thoughts on plausibility in view of the communication of the EBA”

Max Dreisays:

Daniel, I might be completely confused here, but what interests me is the tension between i) the real world of an individual inventor/applicant for patent, and ii) the legal fiction of the “person skilled in the art” whose knowledge of the state of the art is different from that of any one human individual. I think that tension is noticeable when it comes to inspection of the patent application as filed to discern the “problem” of the EPO-PSA obviousness enquiry. Often, there will be a difference between the inventor’s declared problem and an OTP arrived at by reference to a very relevant piece of prior art that was unknown and unconsidered by the drafter of the patent application.

Put another way, what is “plausible” or “speculative” to an imaginary person skilled in the art might be very different from what was for the real inventor. How shall the EPO cope fairly with the reality that the OTP is likely to be different from the subjective problem, however hard the drafter tries to draft in the context of all the close prior art.

Tools for searching the state of the art are ever better, reducing all the time the gap between the skilled person and the real life individuals involved in drafting a patent application. So perhaps in future the drafter will know the OTP before starting to draft. But in the meantime, how shall the EPO-PSA adjudicate plausibility to be fair to real life inventors who don’t yet know the OTP, while deterring them from filing speculatively, and at the same time delivering to the public a level of legal certainty (as to what claims the drafter can take to grant and then enforce) that is “reasonable”. How strongly will the EPO penalise those applicants who didn’t acquire complete knowledge of the state of the art before doing the drafting of their patent application? I look forward with interest to the reasoning of the EBA.


Dear Max Drei,

I would like to draw your attention to R 42 (1,c) which states the following:
“The description shall disclose the invention, as claimed, in such terms that the technical problem, even if not expressly stated as such, and its solution can be understood, and state any advantageous effects of the invention with reference to the background art;”

For instance in a claim for a second medical use, cf. Art 54(5), the effect is a functional feature of the claimed invention. Some claims mention an effect, some do not, hence the distinction between possible objections (Art 83 or 56) made in G 1/03 or T 2001/12.

This is why I do not understand the position of the EBA wanting to limit its considerations to plausibility in the framework of inventive step.

An invention is presumed to be the solution of the technical problem which the inventor faced when he realised his invention, taking into account the background art at his disposal at the time of the invention.

The inventive solution of this technical problem normally brings about some technical advantages or provides a certain effect. Before the PSA, inventive step was affirmed in presence of a surprising effect or a not-expected advantage over the prior art revealed by the search. Most of the time this prior art is closer than the background art of the disposal of the inventor.
The PSA is no more than a tool and can certainly not lead to advantages or effects which were not envisaged by the inventor when he made his invention.

The problem faced by the inventor is not to be confused with the objective technical problem (OTP) resulting from the application of the PSA. Any effect/advantage resulting from the OTP has to lie within the original problem and the advantages/effects the inventor though that his invention brought about.

I can thus not follow your argument there is a “tension between i) the real world of an individual inventor/applicant for patent, and ii) the legal fiction of the “person skilled in the art” whose knowledge of the state of the art is different from that of any one human individual”.

In some applications, it is noticeable that the applicant has better search files than the EPO and it is sometimes difficult to find a better prior art than that disclosed in the application.

To take an image: the original problem is like a Russian doll based on the prior art known to the inventor at the time of his invention. The OTP can only be a smaller doll based on the CPA as it is necessarily closer than that used by the inventor at the time of his invention.

According to the PSA if there no effect discernible with respect of the CPA, the OTP is to provide an alternative.
If it is not alternative, the OTP is to obtain the effect of the invention when starting from the CPA.

Plausibility has thus to be appreciated not with respect of the CPA resulting from the application of the PSA, but vis-á-vis the problem the inventor wanted to solve when he made his invention.

There is thus no reason why the EPO should penalise those applicants who didn’t acquire complete knowledge of the state of the art before doing the drafting of their patent application. I beg to disagree with you on this point.

Max Dreisays:

How interesting, Daniel, at least for me. I was following you until I got to your last two paragraphs. But then, in your penultimate paragraph, you used the words “thus” and “not” and I do not understand how this conclusion follows from your earlier paragraphs. But never mind. If I continue to follow your writings, then one day it will dawn on me, what you are trying to explain to me.

Or perhaps it will all become clear to me when I get to read the G Decision in full.


Dear Max Drei,

Thanks for your comment.

The message I tried to convey, is that the use of the PSA with a closer PA than that of the inventor cannot lead to advantages or effects which were not disclosed, directly or indirectly, in the application as filed.

I am actually not saying anything different as in the position of the president, actually signed by the head of directorate 5.2, Mrs Pihjalamaa.

I refer to point 61. The document is available under the following link:

Max Dreisays:

What a relief, Daniel. Now you are confirming what I was thinking, all along. I shall read with great interest the doc behind your Link. But this only confirms me in my opinion of the value of pre-filing patentability searches.

My point about the benefits of deep pre-filing searches to find the most relevant prior art (the future D1) is that it i) allows the drafter to anticipate how EPO-PSA will play out, ii) enables the drafter to go back to the inventors for a bit more comparative data to include in the application as filed, tuned to D1, to make plausible a specific performance enhancement, one which wasn’t in the first draft and hence iii) improve the prospects of getting dominating claims through to grant.

Acting for an Opponent, one sometimes feels that the patentee got their claims allowed on the basis of comparative data of dubious validity. After grant, it can be hard for an honest opponent to bring the EPO to an opinion that the claims ought not to have been granted. The EPO has to keep permanently in mind the imperative of striking a fair balance of interests, between patent owners and their competitors, reliably upholding good claims and unsentimentally striking down bad ones. In this aspect, I suspect this G Decision will be quite important.


Dear Max Drei,

I fully agree with you that a good pre-filing search is a very important step.

I have been working in fields were it was manifest that some large applicants were having a better search documentation than the EPO had, as there was barely any further PA, besides PA under Art 54(3), to be found.

What I however have discovered is that the present quality of the EPO search is, let’s say, not optimal.

One example is when the opponent comes with PA under Art 54(3) leading to a lack of N of claim 1 as granted-
Other searches are plastered with X documents and the opponent comes with an EP document which kills claim 1 as granted as well as other documents so that the rejection of the opposition is transformed in revocation.
Some searches are only revealing A documents which in general leads to a direct grant (this is what I call the turbo-BEST effect) for the opponent to come with highly relevant documents so that revocation can only be confirmed.

Looked over the three quarters of 2022, in 93,1 % the opponents have brought forward documents which were in the search files of the EPO.
Only in 6,9 % of the cases, the opponents have used documents which could not be available in the EPO search files, like public prior uses, PhD dissertations etc.

That opponents might make more effort to find prior art is to be acknowledged. It is however a matter of fact that there is little non-patent literature used in opposition.

I am not blaming anybody, but simply stating facts. I am continuing my analysis up to the end of the year.

Those overall figures are worrying and can explain recent criticisms of the users of the EPO system as published by JUVE under the following link:

In view of my findings about the quality of search of the EPO, I can only confirm that a thorough pre-filling search is of utmost importance.

As far as plausibility is concerned, I hope that the existing practice will be confirmed, like it was the case for undisclosed disclaimers. I might just add that plausibility has also to do with sufficiency.

Max Dreisays:

Daniel, thanks again for the Link. I must say, having read the President’s “amicus brief” I am left thinking that the referring Board had a blind spot, failed to think the matter through with sufficient rigour, and that whoever wrote the President’s written submission did an excellent job of explaining to the EBA how to arrive at a clear and correct view of the matter.

What’s not to like, in this particular communication “from the desk of the President of the EPO”? In my opinion, nothing. Credit where credit is due, eh?


Dear Max Drei,

The filing of the president’s position by e-mail was quite surprising.
I did nor know that standard e-mail was already a way to communicate with the EBA.

I put this to the striving of EPO’s management to digitise the whole procedure.
This is not as such bad, but it needs the legal framework for it.

Like you, I find the position taken by Directorate 5.2 well argued and not easy to dismiss.

Leave a Reply

Your email address will not be published. Required fields are marked *