On 11.03.2026, the EBA published it provisional opinion on the necessity or not to adapt the description to the claims or vice-versa.
The provisional opinion can be summarised as follows:
Admissibility of the referral
The EBA considered all questions in the referral to be admissible.
Question 1
The EBA made a distinction between two types of inconsistencies
- Those that cause do not cause a non-compliance with the EPС, i.e. “Inconsistency of the First Type”
- Those that cause do cause a non-compliance with the EPС, i.e. “Inconsistency of the Second Type”
The EBA is of the opinion that inconsistencies of the “first type” do not require the adaptation of the description.
The EBA is of the opinion that inconsistencies of the “second type” do require the adaptation of the description. In this situation a further amendment of the description, and/or the claims is necessary in order to comply with the EPC.
Question 1 is to be answered in the positive in case of an inconsistency of the “second type”.
Question 2
The EBA held that the referral decision focuses on Art 84 and distinguished two lines of case law:
- According to the “First Line of Case Law”, Art 84 provides a basis for requiring the adaptation of the description.
- According to the “Second Line of Case Law”, Art 84 does not provide a basis for requiring the adaptation of the description.
For the EBA the “Second Line of Case Law” is represented by: T 1989/18, T 1444/20, T 56/21 and T 2194/19.
The EBA noted that all these cases are ex parte cases, and none pre-date 2021. One of these decisions. T 2194/19 does not seem to diverge from, but agrees, at least in principle, with the “First Line of Case Law”.
Discussion of T 56/21
Turning to T 56/21,the EBA considered that this decision deals the most exhaustively with the issues at hand, in that it adopts a particular interpretation of Art 84.
The board summarised the referral with the following question: “Is the interpretation of Art84 given in T 56/21 correct?”.
The EBA is of the view that many of the conclusions of T 56/21 are inconsistent with decision G 1/24 and its underlying reasoning.
The Enlarged Board is thus of the view that the “First Line of Case Law” should be followed as regards Art 84, and not the approach set out in T 56/21.
Without saying it in so many words, the only answer for Question 2 is unmistakenly clear: the legal basis for requiring the adaptation of the description to the claims or vice versa is Art 84.
Question 3
The EBA sees no reason to draw a distinction between the examination and opposition proceedings in the context of this referral. The answer to Question 3 is “No”.
Comments
In that the EBA considers that T 56/21 is inconsistent with G 1/24, one of the strongest arguments in all the amicus curiae in favour of not adapting the description has fallen into pieces.
When commenting T 56/21 on this blog, I noted that by debating over the relevance of Art 69 in grant proceedings, the present board had actually pre-empted decision G 1/24.
In spite of stating numerous decisions of the German Federal Court (BGH), in T 56/21, the board did not agree that the claims have to be systematically interpreted (Auslegung) before any decision on the validity of the claims and/or on infringement. In G 1/24, the EBA made clear that the description has always to be consulted.
In T 56/21, the board chickened out from a referral, and considered itself able to take a position which it held in conformity with the EPC.
At the time, it would have been wiser for the board to refer the important question of the adaptation of the description to the EBA and not decide ex cathedra. It should have been clear for the board in T 56/21 that the position it took was at odds with that of more or less all the other technical boards.
In T 56/21, the board considered that adapting the description to clear claims would not improve legal certainty. The EBA is manifestly of a divergent opinion.
General conclusion
The legal basis for requiring the adaption the description to the claims or vice versa is beyond any doubt Art 84. After consulting the description according to G 1/24, any inconsistency between claims and description has to be removed by either amending the description or the claims, so that the scope of protection is without any doubt. This is also the position of most of the boards applying G 1/24.
In other words, Art 84 does not only apply from the claims towards the description, but also from the description to the claims.
Comments
43 replies on “G 1/25-Provisional opinion of the EBA-Adaptation of the description can be required”
Thank you – your general conclusion is exactly how I see this question, but it seems to be a rare view among patent attorneys! (And thank you more generally for your blog posts, which I read with great interest and find very helpful. I often say so on LinkedIn, but I don’t think you see it there.)
Dear Anna,
Thanks for your comment. Your position is indeed a rare view among patent attorneys. The clarification thought for in G 1/25 is for the benefit of third parties. Those of your colleagues vehemently again adaptation of the description to the claims and vice-versa, seem to forget an important point: in some situations they might represent the proprietor, in other situations, they might represent an opponent and/or a potential infringer. Clarifying the sope of protection is thus beneficial for both sides.
Thanks also for your kind words about my blog. I will certainly continue. I used to be on LinkedIn, but a few years ago, I decided to withdraw from any social media. Furthermore, LinkedIn does contain too much “noise”, i.e. communications which have little to do with the professional life of members, drifting toward political topics, which have nothing to do with the professional life of the members.
Thanks also for your kind words about my blog. I used to be on LinkedIn, but a few years ago, I decided to withdraw from any social media. Furthermore, LinkedIn does contain too much “noise”, i.e. communications which have little to do with the professional life of members.
A further reason for leaving LinkedIn was the comment of many readers at the time, that it was rather difficult to find my “blog entries” in the host of information published on LinkedIn. When IP.appify offered me the possibility to publish on their blog, I took the opportunity, as it avoided me having to start my own Internet page. Now any new publication on IPappify Blog is announced in LinkedIn.
Daniel
I can’t get my head around it: Inconsistencies that do not cause a non-compliance with the EPC, do not require the adaptation of the description, and those that cause do cause a non-compliance with the EPC, do require the adaptation of the description.
Of course that’s obvious, nobody can object to that statement.
But this is just moving the key issue somewhere else. How are users to identify which inconsistencies do, or do not, cause a non-compliance with the EPC?
The argument between applicant and ED/OD will shift from ‘does my inconsistency require amendment of the description?’ to ‘is my inconsistency of the first type or of the second type?’
It can still be argued (as much as before) that any ‘inconsistency’ that does not explicitly contradict the claims is an inconsistency of the first type and therefore does not require an adaptation of the description.
So really, I don’t see how this opinion clarifies the law.
@bj,
You should not forget that what will clarify the law, is not the provisional opinion of the EBA, it is the actual decision and the answer of the EBA to the questions raised.
The only derivable conclusion is that the EBA appears to be of the opinion that the description will have to be adapted to the claims AND/OR the claims to the description, be it in opposition, examination or in appeal, and that the legal basis for this is Art 84.
It is your perfect right to argue that an inconsistency between your claims and your description is one of the first type. Indeed, any ‘inconsistency’ that does not explicitly contradict the claims is an inconsistency of the first type.
At the end of the day, it will be to the ED/OD and a TBA to decide which type of inconsistency is at stake.
To me, any inconsistency giving rise to a different interpretation of the plain wording of the claim compared to that derivable from the description, the latter being broadening or limiting, will be objected under Art 84, as the scope of protection is then unclear.
Another way to look at it: leaving the possibility of different interpretations between claims and description and vice-versa, does not represent a clear contradiction between claims and description, but will end up with an objection under Art 84, as the scope of protection is then unclear.
I know this is only a preliminary opinion, and my comment was based on the hypothetical situation that this preliminary opinion will be maintained.
What struck me in relation to Q1, is that it seems to be merely stating the obvious:
Question: ‘If I have an inconsistency, do I need to adapt the description in order to comply with the law?’
Preliminary Answer: ‘Only if the inconsistency is such that it causes a non-compliance with the law, otherwise not’
This is not clarifying a point of law. But hopefully more detail will come in the final decision.
@bj,
I look at the provisional opinion of the EBA like a document set up by legally qualified people.
The EBA defined firstly two types of inconsistencies and then secondly two types of case law when it comes to the interpretation of Art 84.
The presentation of those four possibilities is a standard way of drafting a legal analysis of a situation, when a plurality of possibilities can be distinguished.
It indeed looks like stating the obvious, but the contrast between the two types inconsistencies eventually clarifies the situation. Only those which imply a non-conformity with the EPC are to be taken into account.
For instance, a misspelling of the same word in the claim and in the description is irrelevant. It can be rectified as a typo under R 139 in procedures before the EPO and under R 140 if the typo is in the decision after it has been issued.
Hello Daniel,
Thank you for your clarification. I know you and I have a different opinion on this matter, but this is not an attempt to relight that fire.
You wrote ‘It indeed looks like stating the obvious, but the contrast between the two types inconsistencies eventually clarifies the situation. Only those which imply a non-conformity with the EPC are to be taken into account. ‘
To me, inconsistencies that do not cause non-compliance with the EPC were never even part of the referred question, there were never any disagreements about such inconsistencies.
I agree, bj. The hotly debated question is “can X ever be a problem”. The EBA has said “X can be divided into two logical classes; X which is a problem and X which is not a problem. X which is a problem is indeed a problem.” This is, even by the EBA’s standards, a remarkably trite and unhelpful answer. I wonder if there are split views within the EBA. If there are then it seems this could be their attempt to come to a fudge which is uncontroversial and which does not upset anyone.
Dear Daniel X Thomas, thanks for your comments on this subject-matter. To me, the EBA’s message is clear: if the applicant has amended the claims to their likings and this amended claim is not supported by the description (w.r.t. Art. 84 EPC), there shall be an amendment. And since the claims have been amended to overcome some sort of objection, the EBA’s amendmant shall take place in the description.
What I do not understand fully, and maybe you can shed some light on this, is what kind of “Inconsistency of the Second Type” can occur? Obviously, there might be a problem with Art. 84. But what about inconsitencies that arise due to the application of G1/24? Would this also count as a “non-compliance with the EPC”?
Dear Arjun P,
Thanks for your comment.
I take from the provisional opinion of the EBA, that any inconsistency giving rise to doubts about the scope of protection has to lead to an amendment of the claim, but the EBA has clearly indicated that vice-versa, the claim can also be adapted to the description. This means that Art 84 applies in both directions, from the claims to the description, but also from the description towards the claims. It is left to a party to decide which direction it choses, cf. party disposition, under Art 114(2).
A good example is that in the Agfa/Gucci case. The original claim covered chromatic+achromatic. In view of the prior art, the independent claim was limited to achromatic, but chromatic was not deleted from the description.
Another example is when an independent claim is limited by a feature qualified in the description as being optional. Once the feature is used to limit the independent claim, it cannot be left as optional in the description.
In both those cases, not adapting the description gives rise to doubt about the scope of protection.
Reasons 20 of G 1/24 read:
“The above considerations highlight the importance of the ED carrying out a high quality examination of whether a claim fulfils the clarity requirements of Art 84. The correct response to any unclarity in a claim is amendment. This approach was emphasised in the Comments of the President of the EPO”.
When looking at the case law of the boards following G 1/24, one can see that most of the boards are of the opinion that, should the interpretation resulting from the plain wording of the claims differ from that possible from the description, the latter being broadening or limiting, if the applicant/proprietor wishes the interpretation from the description to be taken into account, it has to amend the claim accordingly when it comes to interpretation of the claims under Art 52-57. .
When it comes to limiting features in the description, see for example: T 1232/23, T 1465/23, T 1846/23, T 1069/23, T 1351/23, T 2027/23, T 412/24, T 2034/23, T 1402/24.
When it comes to broadening features in the description, see for instance T 1849/23 and T 439/22. The latter does however depart from the approach of all the other boards as it falls back in a holistic interpretation of the claim, in that it holds that any interpretation found in the description should be read in the claim, whether broadening or limiting. I doubt that this is the result thought of by the EBA in G 1/24. I take from G 1/24 and probably G 1/25, that any inconsistency between claims and description has to be queried under Art 84.
One decision is particularly clear: T 2027/23: it is not the task of the boards of appeal to reach alignment between claims and description by way of interpretative somersaults. The board added that by allowing a misalignment between claims and description, it would allow the applicant/proprietor to nurture an Angora cat.
@ bj and anonymous,
I can understand bj’s point of view, and bj is entitled to have a different opinion on this matter. I do also not consider bj’s comment as “an attempt to relight that fire”.
It might well be that the question raised could for the board be approached in a systematic way as it did. I do not see anything wrong with this. The EBA is free to approach the questions referred to it as it thinks fit. Some might disagree with the path chosen, but nobody can dictate the EBA the way to come to its conclusion. This should not be forgotten.
Coming with speculations and wondering if there are split views within the EBA, is far-fetched. I find to the contrary that in view of the clear statements of the EBA, albeit provisional, we have actually a proof of the contrary.
The criticism of T 56/21 in the provisional opinion is particularly strong. In T 56/21, the board was vehemently and adamantly of the opinion needs not be adapted, is rather a proof. I think that the EBA has been telling off the board of T 56/21 for its interpretation of Art 84.
Most, if not all amicus curiae against adaptation of the description relied on T 56/21. It is clear that the EBA is not inclined to follow the position adopted in T 56/21. Most of the amicus curiae against adaptation of the description have alleged that there is no basis in the EPC for requiring the adaptation of the description. They have clearly been proven wrong.
Even provisional, the opinion of the EBA is abundantly clear, and I have strong doubts that at the end of the day, the EBA will do anything avoiding to come to “a fudge which is uncontroversial and which does not upset anyone”.
The EBA has really gone a long way out, and it is not to be expected that it will change its position in a fundamental manner.
Assume that an independent claim is limited in opposition by introduction of features from a dependent claim. How then would A84 be relied upon to enforce an adaptation of the description, bearing in mind G 3/14?
@ Curious George,
In the executive summary of my amicus curiae for G 1/25, I wrote:
“Question 1 is to be answered positively, under the proviso of G 3/14, but this proviso could be lifted”.
The title of Section III.1.5.6.2 reads: “Limits of the adaptation of the description in opposition and/or opposition appeal procedures”
In this section I drew the attention to the EBA that there might be situations in which adaptation of the description does not appear be possible in view of G 3/14.
In reply to Question 3, I wrote:
“In examination and examination appeal procedures, there is no limitation under G 3/14 as in opposition and opposition appeal procedures”.
When reading attentively my amicus curiae brief for G 1/25, you would have realised that I was aware of the problem. Your question did not surprise me. I thought of it a while ago.
It will be interesting to see how the EBA deals with the problem.
Thank you Mr Thomas for all your posts on this topic, particularly the summary of all of the submissions from different parties.
A very frustrating thing about all of this is that, while people can say that adaptation of the description has always been required by the EPC, the simple matter is that this has not been applied at the level currently being decided on (i.e. having to delete/disclaim all embodiments falling outside the scope of the claims and make amendments to remove optional terminology in relation to independent claim features) to the vast majority of European patent applications since the EPC first came into effect.
I could go on EspaceNet now and pick 100 European patents at random, and I would guarantee that 90+% have merely had description amendments to the effect of changes to the Summary of Invention section.
As this recent ‘description adaption’ affair has allegedly been about providing “high quality patents” (according to those in favour, including the EPO’s president), one can only conclude that this means all of those older patents are not high quality. I hope the EPO plan on giving some kind of compensation to past applicants for the apparently shoddy examination job that was done on those older cases; at least 50% of all renewals fees paid during prosecution of each application would do it. This might go some way towards balancing out the increased costs that applicants are going to be lumped with in the future; or some of it could go towards environmental schemes to offset the forthcoming use of AI in preparing amended descriptions, because it is an expense that many applicants cannot bear and neither can attorney firms subsidise it.
It’s amazing that national courts of contracting states have been able to handle litigation involving validated EPs until now, considering how terribly confusing it apparently is to have a description including an embodiment which is not within the claims.
Let’s see what the final decision says. With enough pressure from industry, maybe we’ll see an EPC2026 which redrafts Art 84 to say “The claims shall define the matter for which protection is sought. They shall be clear and concise and be supported by the description to the extent that the description includes at least one embodiment within the scope of the claims.”
@ Frustrated Attorney,
I can fully understand your frustration.
I agree with you that the adaptation of the description has not been handled in an uniform way by EDs and ODs. One explanation is that there have been clear instructions by superiors, of course only orally and not in writing, that, since Art 84 is not a ground for opposition, clarity does not really matter. If the examiner is able to understand the claim, it is good enough, and reading the description is actually superfluous, be it before the search or before issuing the communication under R 71(3). It has never been my position while in charge of examiners.
The attitude is that, if the independent claim is novel and inventive, then it is not worth to look at the dependent claims. This is only correct for the dependent claims as far as novelty and inventive step are concerned. It does not say anything about the clarity of dependent claims as such and in combination with the independent claim they are attached to, or other dependent claims. It does also not say anything about added matter in a dependent claim or as result of the reordering of claims. In opposition, or when consulted, I have been faced with patents for which I wondered how they could have reached the grant stage, epecially when Art 84 comes into play.
I am sure that there are lots of patents granted in which the description has not been adapted to the claims. The best example is the famous Agfa/Gucci case. Whilst the claim had been limited to achromatic, the variant chromatic was still mentioned in the description. Even the CFI-LD Hamburg complained that the description had not been adapted and every reference to chromatic deleted.
The “recent ‘description adaption’ affair has allegedly been about providing “high quality patents” (according to those in favour, including the EPO’s president)”, but the insistence n the adaptation of the description has, in my opinion, been triggered by the famous “pemetrexed disodium case”.
I fear that your hope that the EPO plans on giving some kind of compensation to past applicants for the apparently shoddy examination job that was done on those older cases, whereby at least 50% of all renewals fees paid during prosecution of each application would do it, will remain a hope which will never become reality, be it only for the fact that there is no legal basis for such a compensation. All the renewal fees cashed in by the EPO during prosecution have been gone for a long time.
It is not amazing for me that national courts of contracting states have been able to handle litigation involving validated EPs until now, considering how terribly confusing it apparently is to have a description including an embodiment which is not within the claims. Its their job when faced with nullity or infringing action.
The problem is due to the fact that patents ending in court are a small number compared to all patents granted every year by the EPO. When an examiner gets an application on its desk, he does not know whether this patent will be the pearl in the oyster or will end up in the Bermuda triangle of patents not having reached the market place. Therefore, all applications should receive the same attention.
As far as your wish to see Art 84 amended, I do not see it coming to reality in 2026 or in the near or even distant future. Amending the EPC needs a Diplomatic Conference. When you see that Art 4a EPC 2000 was dead borne, I cannot see any amendment of Art 84. At the end of the EPC 2000 conference, there was still a “second basket” with amendments of substance to the EPC, but there is nothing forthcoming at this level.
An amendment of the Implementing Rules is also a no go in view of Art 164(1). The only possibility would be a dynamic interpretation of the EPC, but the inventor of the present EBA is part of the panel of the EBA but not rapporteur as in G 3/19.
@Frustrated Attorney
The EPO’s practice regarding the adaptation of the description appears to have completely changed since (i) the reminder to Examining Divisions that they are not entitled to enter amendments at the R 71(3) stage without the applicant’s agreement (and this is a serious quality issue, judging from the EPO quality report), and (ii) the 2024 revision of Guideline F-IV 4.3, which now sets out that an inconsistency violates the clarity requirement of Article 84 if it “casts doubt” (in the present tense) as to the meaning of the claim, instead of the hypothetical “could cast doubt” in the 2021 version. This clearly places the burden of proof on the Examining Division to show by specific reasoning an inconsistency in violation of the EPC. According to feedback from colleagues, the practice of amendments entered by examining divisions at the R 71(3) stage without the applicant’s agreement seems to have disappeared. This confirms your own observation.
Keeping this in mind, several positive points can be found in the EBA’s provisional opinion.
– The emphasis on determining what constitutes an inconsistency contrary to the EPC supports the revised wording of the directive. It is of note tat this wording is in line with the PCT ISPE guidelines 5.29.
– The emphasis on Article 84, which disregards political arguments not based on the EPC, such as the claim to assign the EPO a role as guardian of the public interest, implying interference with the proceedings before the courts. In addition, the concern for legal certainty supporting this claim is at odds with the legal uncertainty caused by Articles 123(2) and/or 123(3) issues arising from the adaptation of the description.
– The emphasis on Article 84 contains no explicit justification for imposing an adaptation of the description. If the Examining Division asserts the existence of an inconsistency in violation of Article 84 in the form of a communication, it is then up to the applicant, and the applicant alone, to decide how to remedy it, if they admit the existence of such an inconsistency, in accordance with the principle of Article 113(2). And it is often much simpler and safer to cure the inconsistency by amending the claim than the description.
fh: it is encouraging to read your comment but I am not as sanguine as you about what will happen, going forwards, when ED’s examine for compliance with Art 84, EPC.
What I expect is that it will be easy for an EPO Examiner to identify an inconsistency and even more easy then to object that the inconsistency renders less than 100% clear the scope of protection of the claim under examination. Even after “consulting” the description to fix the context under which each word of the claim is presented, the meaning of any given sentence (at least when in the English language) is hardly ever 100% “clear” and if it is 95% clear then is the claim “clear” or “unclear”?
Proving a negative is tough. When the ED avers that my claim is less than 100% clear, how shall I prove that it is not, to this extent, unclear?
Max Drei,
In theory I can follow your line of reasoning, but it is all hypothetical. In most of the cases it will be manifest when an inconsistency between claims and description or vice-versa occurs. Refusing the adaptation of the description because of such theoretical considerations is simply a no go. Let’s stay reasonable.
Mr Hagel,
It should never have come as far as it is went, so that ED’s are twisting the arm of the applicant to bring in amendments of substance at the R 71(3) stage. This is due to the ever growing pressure put on examiners to grant as quickly, as much as possible, in the less time possible. In any case, it was not, and still is not, correct.
It is nothing new under the sun that any amendment requested by an ED/OD has to be justified, and not a legal rule merely thrown at the applicant/proprietor, be it Art 84, 54, 56, 83, 123(2] or 123(3). This is a problem when the ED amends the description at the R 71(3) stage, and merely writes Art 84 without any justification.
You are playing with words by distinguishing between it “casts doubt” compared to “could cast doubt”. In lots of cases, there is no adaptation of the description at all in spite of the necessity of it according to the valid Guidelines. Just look at the case with chromatic left in the description, although the claim was limited to achromatic.
As Art 84 is not a ground of opposition, it was, and most probably still is, common practice in certain technical areas to never read the description and certainly not to check if an adaptation of the description was actually caried out by the applicant when the claims were limited. This should change with G 1/24.
It is well known that you are a great fan of the PCT Guidelines. At the end of the day, what matters is not what the PCT Guidelines say, but what the EPO Guidelines say. This is the more so since the EPO is not even a contracting state of the PCT, although it acts as ISA and IPEA. Only when the EPO acts in such a role, it is bound by the PCT Guidelines. In the regional phase, the EPO Guidelines apply and the PCT Guidelines become irrelevant.
Besides one case in the UK, where the applicant amended the description on its own volition, and thereby ended up with added matter, I am still waiting to see one case in which a patent has been revoked by a national court or the UPC due to the adaptation of the description to the claims or vice versa. If done properly, the adaptation of the description will not induce any problems under Art 123(2) or 123(3).
I fully agree and support the principle of party disposition enshrined in Art 113(2). Faced with an objection under Art 84, it is indeed for the applicant to propose a text, and not for an ED to force a certain wording, but at the end of the day, it will be the ED to decide whether the amendment of the description or of the claims is in accordance with the EPC or not. If its end in the refusal of the application or the revocation of the patent due to an incorrect adaptation of the description, the applicant/proprietor can always file an appeal. In opposition, Art 84 is only applicable within the limits of G 3/14.
As member/chair of numerous ODs, I have seen cases in which, after remittal, proprietor and opponent haggled for many hours about the adaptation of the description. In lots of cases, the opponent wanted a certain adaptation of the description so that it could circumvent the patent in an easier way. It also happened that it was the proprietor wishing an adaptation of the description, which made it easier for him to catch a potential infringer. At the end of the day, it is the OD to decide what was allowable or not and not the parties.
Whether it is the claim which is adapted to the description or the other way round is actually irrelevant. Any inconsistency/discrepancy between claims and description allowing the applicant/proprietor to change the interpretation as he thinks fit, is to be removed for the sake of legal certainty of third parties.
Mr Thomas,
There is a major difference between « could cast doubt » and « casts doubt ». This is not just playing with words.
The « casts doubt » language is in the present tense and the burden of proof is on the examiner to show that the inconsistency indeed casts doubt. In the practice of description amendments entered at the R71(3) stage without prior consultation with the applicant, it was common for examiners to provide zero specific reasoning to support the entry, other than a mere reference to the guideline. I can see that you disapprove such practice.
Amendments consisting of deletions or disclaiming statements regarding embodiments which are no longer covered by the claims have no reason to be always equated to inconsistencies which « cast doubt ». The examiner may think this is the case, but then the onus is on the examiner to provide specific reasoning for this position. And the specific reasoning is a necessary basis for an applicant minded to challenge the examiner’s objection.
I have smiled at your assertion that I am well known as a great fun of the PCT guidelines. When you say I am « well known » for that, it is rather flattering, thank you for this kind comment. But honestly, it is overstated, to say the least. And when you say I am a great fun of the PCT guidelines, this is true for the one dealing with inconsistencies. My reasons are that first, I find the PCT guideline provides a reasonable, nuanced approach. Second, the PCT guideline reflects a broad international consensus. Third, it make sense from a user perspective for the EPO and PCT guidelines to be substantially aligned.
@MaxDrei
You are correct that I am sanguine. I am this way in general, I tend to see a glass half-full when some people see it half-empty. On the clarity issue though, I have an objective reason for being sanguine. If the onus is on the examiner to prove that an inconsistency casts doubt, it is not sufficient to simply assert a claim is less than 100% clear (I doubt there is ever absolute clarity anyway). The examiner has to provide specific reasoning, Then you do not have to prove a negative. You have a basis that you can challenge by specific rebuttal arguments.
All very helpful, thanks FH. BUT: onus here, onus there, what caught my eye in DXT’s comment is his “many hours of haggling” over the issue whether any given schedule of conforming amendment is or is not EPC-compliant.
Presumably there was more in these hours of attorney argument than an endless childish repetition of an “It is. No it isn’t”. Presumably, patent owner and opponent tried every reasoned line of argument they could generate. And yet the debate still went on for hours. Why should it be otherwise between Applicant and ED, whenever Aplicant is given a fair opportunity to reply.
DXT chides me for being over-theoretical. But when I argue from a specific hypothetical, a specific matrix of fact, I am chided for arguing from the particular to the general. Never mind. Daniel’s confirmation that a requirement for strict conformity often results in many hours of haggling is much appreciated.
Max Drei,
Again let’s be reasonable. There is no need to haggle for hours between an ED and an applicant. Contrary to the proprietor and/or the opponent, the ED has no vested interests in the outcome of the case.
The ED is there to guarantee that the applicant/proprietor cannot change its interpretation of the claim as he thinks fit, by leaving in the description statements allowing to cast doubts on the matter for which protection is thought. In presence of such an inconsistency, the matter for which protection is thought is manifestly ill defined and Art 84 applies.
Once all the arguments have been exchanged, and the right to be heard has been respected, it is for the ED to take an appealable decision. There is no need for lengthy discussions as between opponent and proprietor.
The two situations are not comparable at all. In opposition, it is for the OD to decide what it holds as acceptable and not for the parties to decide. The OD is then to issue an appealable decision. In none of the cases in which I was involved, a party to opposition appealed an OD’s decision defining the adaptation of the description it held to be correct and acceptable, even after an unnecessary lengthy haggling.
An OD cannot bluntly stop the unnecessary haggling as it otherwise could commit a SPV. The impression gained in such situations was rather that the respective representatives were in search of billable hours. But this is not what is relevant for an ED or an OD.
Mr Hagel,
Whether there is a major difference between “could cast doubt” and “casts doubt”, is actually irrelevant, as in any case you added rightly that the burden of proof of an inconsistency is on the side of the ED. It is an ascertained fact that an ED cannot, in the communication under R 71(3) merely throw at the applicant Art 84 without any explanation. No need to dwell on it.
I strongly disagree with you, that amendments consisting of disclaiming statements regarding embodiments which are no longer covered by the claims are not to be always equated to inconsistencies which “cast doubt”.
My position has always been that, in principle nothing ought to be deleted in case of a plurality of more than n=2 embodiments, besides embodiments not any longer falling under the claims. Those have simply to be marked as such. When n=3, and if only embodiment 1 falls under the claim, embodiments 2 and 3 cannot be left as representing the claimed invention. When n=3, and two embodiments still fall under the claim, then the remaining one cannot be left as representing the claimed invention.
In the presence of only two embodiments, whereby one is not any longer covered by the claims, deletion is however necessary. The same applies when an optional feature in the description has been used in limiting an independent claim. Then, the optional character in the description has to be deleted before grant or maintenance in amended form.
In all those cases, the matter for which protection is thought is manifestly ill defined and Art 84 applies.
It will avoid the proprietor to claim later, that although embodiments that are not any longer falling under the wording of the claim, might still be held valid for infringement. I repeat once more the example of chromatic and achromatic, which you carefully dispensed yourself from commenting.
Contrary to what you think, the EPO Guidelines also provide a reasonable, nuanced approach. That those are not always correctly applied is a different matter, but it does not warrant to cling to the PCT Guidelines. The latter can as well be ill applied and then we are in the same situation than at the EPO.
The EPO acts as ISA and IPEA and a lot of what is happening under Chapter I and II of the PCT has been inspired by the EPO. However, when an application enters the regional phase before the EPO, the EPC applies and EDs are bound by the EPO Guidelines and not by those of the PCT. After all, the PCT does not end up with a granted patent. This decision is left to national or regional patent offices. This is a major difference between the PCT and the EPC.
Here I would think a reference to Article 27(1) PCT is appropriate. No national law shall require compliance with requirements relating to the form or contents of the international application different from or
additional to those which are provided for in this Treaty and the
Regulations.
@ Pat
Your reference to Article 27(1) PCT is interesting.
I do however allow myself to observe that Article 6 PCT says:
“The claim or claims shall define the matter for which protection is sought. Claims shall be clear and concise. They shall be fully supported by the description.”
Art 84 EPC says:
“The claims shall define the matter for which protection is sought. They shall be clear and concise and be supported by the description.”
Do you spot the difference?
As the EPC requires less than what the PCT requires, the EPO is clearly acting within the boundaries of the PCT as set by Art 27(1) PCT.
I allow myself to conclude that your comment somehow backfired.
The “Travaux Préparatoires” to Art 84 EPC do not say much about it. The word “fully” was deleted at the request of associations of representatives.
Some representatives and associations of those therefore conclude that Art 84 only applies from the claims to the description, and not the other way round.
I am not at all convinced by this argumentation and I have been trained to believe that it also applies the other way round. This is also what I informed the examiners I was in charge of.
This positions appears to reflected in the provisional opinion of the EBA when it concludes that, in case of an inconsistency of the second type, “a further amendment of the description, and/or the claims is necessary in order to comply with the EPC”.
In my humble opinion, the situation is thereby crystal clear.
Unless some really compelling arguments come from the opponent and the proprietor, one can guess the outcome of the referral.
Would anyone be able to help shed some light on the latest submission by the opponent and the TPO? To my humble understanding, neither party to the proceedings, neither the patentee nor the opponent, believes that Article 84 is the basis for the amendment of the specification. I wonder how this squares with the approach taken by the EBA. Furthermore, it seems that a TPO has been filed challenging the composition of the EBA. It would be great to understand if there is anything to make of it.
@ Anonymous,
Whilst unsigned TPO have been admitted in examination appeal proceedings, this has never been the case in opposition appeal proceedings. When looking briefly at the case law of the boards in matters of TPO in appeal, even TPO being clearly identifiable those have little or no chance to be admitted once the board has summoned to OP.
As far as anonymous OP are filed at the opposition appeal stage, they are not admitted, as an anonymous TPO would accord the third party a more favourable treatment than would be given to the parties. It is also not sure that the TPO have not been filed by one of the parties in order to circumvent the limitations imposed by the RPBA to the parties.
It is appears thus more likely than not, that the EBA will disregard the TPO, even if one of the parties would use the argumentation in the TPO during OP. Such an argumnt would be late filed.
As far as partiality is concerned, only a party to the procedure, or the members of the EBA can raise a objection of partiality against one of its members. This happened in G 1/21, where the chair of the EBA has been removed from the panel by its peers. This is unlikely to happen in the present case.
The style and the objections raised in the TPO rings a bell, as similar objections have been raised in the IPKat post “BREAKING: Imminent new referral to the EBA on the relevance of G1/24 to added matter” of 03.03.2026.
Here is the link:
https://ipkitten.blogspot.com/2026/03/breaking-imminent-new-referral-to-eba.html
Just have a look at the objections against the provisional opinion of the EBA stemming from “Proof of the pudding” from 16.03.2026 at 11:07:00 GMT onwards. As far as suspected partiality is concerned look at the comment of Proof of the Pudding of 23.03.2026 at 11:04:00 GMT. The resemblance between the points raised in the anonymous TPO and the statements of Proof of the pudding in IPKat is striking.
Proof of the pudding and most of the amicus curiae rely heavily on T 56/21 which has been found at odds with G 1/24 by the EBA. One of the key arguments of those advocating that the description does not have to be adapted lose thereby their best argument. In its latest submission of 01.04.2026, the opponent is trying desperately to hold on to T 56/21, see section “3. Considerations re. T 56/21 and G 1/24”.
The opponent also repeats the litany of reasons why the description should not be adapted. It is not by repeating statements that they become compelling.
I invite you to read Christoph D. Schober’s article published in epi information 1/26:
Aligning Claims and Description – Why G 1/25 Matters for Legal Certainty
Available at
https://information.patentepi.org/issue-1-2026/why-g-1-25-matters-for-legal-certainty.html
I have not often read such a passionate plea from a European patent attorney in favour of adapting the description to the claims (or vice versa). This clearly contradicts the opponent’s statement in section of its latest submission: “4. Legal certainty for third parties”
I make no secret of the fact that I agree with the views expressed in this article.
This article is a clear counterpoint to the article by Ms Schäfer and Mr Majdic published in epi information 4/25:
Anpassung der Beschreibung durch Streichung – (k)eine bloße Formsache?
Available at
https://information.patentepi.org/issue-4-2025/anpassung-der-beschreibung-durch-streichung.html
You are absolutely right that TPOs are seldom taken into account in opposition proceedings. G1/03 for example explicitly mentions that anonymous TPOs were “[not] taken into account by the Enlarged Board.” However, this is up to the board to decide and there is room for discretion (but rightly not so for the partiallity argument raised therein. This is clearly up to the parties or the individual member to raise. I mean, this already appears in the Article itself).
I still find one argument (at least how I interpret the TPO) interesting, namely that they should engage with T56/21 on an intellectual level instead of the seemingly formalistic approach. I mean in the EQE one is also taught to not only attack an object once in case of partial priority. In principle a claim is dead if this sole attack succeeds. The proprietor has to react and amend the claim, thus leaving open traps for Art. 123. So, we are teaching candidates to think further and address all issues of a claim, but it appears the board didn’t do so in the preliminary opinion, because T56/21 deviates only in one detail that is incompatible with G1/24, while most of the reasonings appear, at least to me, to be compatible with G1/24. And they didn’t even address the reasonings in the referring decision or in the amicus.
@ BF
You might not be aware of it, but I have been training people for papers A-C of the EQE for over 20 years. I stopped now in view of the new EQE.
In real life, as opponent, you have to file as many attacks as possible against the patent you oppose. This was also the case in the old paper C of the EQE. The papers were so conceived that a plurality of attack were sometimes possible for a claim, be it independent or dependent.
At the end of the day, if the OD or a board finds one reason to find a claim not allowable, be it independent or dependent, the whole set of claims is not allowable. There is no need to go through all objections. An OD has never been sanctioned by a board for committing a SPV if it has not dealt with all objections raised against a claim. It was enough if the reason decided by the OD was compelling.
It is clear that any claim amendment brings with it the risk of added matter. This can be the case for the independent claim as such, but also for the combination of the new independent claim with the existing dependent claims, cf. old paper B of the EQE. It is primarily the applicant’s/proprietor’s responsibility. An ED/OD is only there to check whether the requirements of the EPC are met or not.
In my opinion, you are unfairly criticising the EBA by asking it to set out its full reasoning regarding T 56/21. Since most of the submissions wanting the reply NO to question 1, rely heavily on T 56/21, it was to be expected that the EBA would say something about this decision in its provisional opinion. Requesting a complete argumentation is not what should be subject-matter of a provisional opinion.
It can be agree with T 56/21 that Art 69 has nothing to do in examination proceedings. As far as opposition proceedings are concerned, the scope of protection is only to be checked under Art 123(3). However, a broader definition of the matter for which protection is thought in a claim is a no go in opposition, even if the latter was contained in the original description. It can also be agreed with T 56/21 that examination of Art 84 not linked to examination of Art 69. The agreement stops there.
Nevertheless, where T 56/21 is in clear contradiction with G 1/24 is when it alleges that. if claims of more limited subject-matter are deemed allowable, any statement in the description allowing a broader definition of the matter for which protection is thought must be ignored. This means that the description does not have to be consulted.
This is manifestly in contradiction with the order in G 1/24. Even this was queried in T 439/22.
Moreover, keeping in the description a broader definition of the matter for which protection is thought, is clearly inviting the proprietor to come up with a nice Agora cat in post-grant or post-opposition proceedings.
The EBA has certainly the discretion to admit or not anonymous TPO, but I doubt that this will occur in the present instance. If the EBA would do so, it would be more an exception than the rule.
As far as the anonymous TPO is concerned, I intend to deal with it in a separate blog post. It is not only aggressive in tone and contains severe accusations of partiality, which are not at all justified, but it is not compelling whatsoever.
The anonymous TPO content will certainly not render the EBA lenient so as to take it into account. It might simply disregarded it, and not at all take position.
There is another TPO by Michael Snodin. It is actually very interesting as well, because it directly addresses what you apparently deem as decisive regarding incompatibility between T56/21 and G1/24: The first line of case law isn’t compatible either, because these decisions, at least the ones talking about claim interpretation, actually agree with T56/21 that the description is to be disregarded for claim interpretation. And due to G1/24 the angora cat problem isn’t there anymore, because the description now always has to be consulted and broadening has to be taken into account, even if this does not apply to limitations (even though I believe we might see another referral regarding this, now that T439/22 is out).
And this is why I actually agree with the anonymous TPO: they should address the legal reasoning in T56/21 at least. At least from Reasoning 53 onwards in T56/21, because there the reasoning doesn’t rely on what seems to be incompatible with G1/24 anymore, but solely on the literal wording of Art. 84 EPC and on the link between Art 69 and 84 EPC assumed by several Boards and given as a reason for requiring description amendments.
A propos the Schober article, I have to say that I find it shallow and deeply disappointing. Yes, of course the goal of “legal certainty” is important enough to trump almost every other consideration. But given the difference between what is defined in the claim as the subject matter being protected and what Art 69 EPC defines as the scope of protection, I do not see the simplistic goal of “legal certainty” as achieved when there is strict conformity but not achieved as soon as an “inconsistency” can be identified, between the description and the claim. On the contrary, I can imagine that an egregious degree of prosecution amendment of the description during prosecution could complicate rather than simplify the assessment (in contested inter Partes infringement and validity litigation) under Art 69 EPC of what exactly is the correct “scope of protection”.
From a judge of the UPC I would have expected more persuasive reasoning, as opposed to a pious declaration that absent strict conformity legal certainty is unobtainable. I hope for more persuasive reasoning from the EBA.
Max Drei,
That you do not agree with the Schober article does not come as a surprise as you are clearly against alignment of claims and description.
It as well is possible to consider your comment as shallow and deeply disappointing. I expect from your part something more compelling than your comment.
I leave you the responsibility for what you said about the UPC judge. He has put in nutshell what actually matters and cannot be blamed for it.
I repeat once more that Art 84 and R 42 (1,c) relate to the definition of the matter for which protection is sought. Art 69 defines the scope of protection, which is something different as it can include equivalents. At grant and/or after maintenance in amended form, it is the duty of the EPO to insure that the definitions of the matter for which protection is thought is identical in the description or in the claim.
Legal certainty is not a mere simplistic goal. It means primarily that the description cannot contain a definition of the matter for which protection is sought is broader or more limited than that which is derivable from the plain wording of the claim. This has nothing at all to do with an egregious degree of prosecution amendment of the description during prosecution.
I am not for generally deleting parts of the original disclosure, but embodiments that do not any longer fall under the definition in the claim of the matter for which protection is sought should be marked accordingly. What has to be deleted is the optional character in the description of a feature which has been introduced as limitation in an independent claim. This is not asking too much.
It is certainly debatable whether 100% legal certainty or 100% clarity is at all attainable. At least prima facie contradictions between definitions in the claim and definitions in the description have to be avoided.
This is why there is Art 69 for post grant proceedings. This is the domain of national courts and of the UPC. On the basis of a patent in which description and claims are aligned those jurisdictions are free to decide what they consider being the actual scope of protection.
That they can also take into account, or not, prosecution history is for them to decide and is out of the reach of the EPO. At least the EPO should grant or maintain patents for there is no haggling possible around the definition of the matter for which protection is sought. In other words, the applicant/proprietor should not be in a position to choose the definition which suits him best depending on the situation.
As of now, for Examiners at the EPO, detailed description of devices and methods that fall outside the literal wording of the claim has to be deleted or tagged with words that declare that device or method to be “not the invention”. But wait! Daniel, you now write that they have to be “marked accordingly”. Specifically, what if these passages describe subject matter that is plainly equivalent to what is literally claimed. Is it fair to the inventor, to require as a condition of acceptance and grant, that these embodiments, within the scope of protection, be stated explicitly to be “not” the invention? I mean, in the context of the EPC, Art 84 and 69, what is “the invention”?
Perhaps with your “marked accordingly”, Daniel, you are aware of the issue and have in mind some form of words other than “not the invention”? (I realise of course that, in any real case, the EPO is not able to put words in the mouth of the applicant).
It could be, Daniel, that there is not as much between us as some of the comments in this thread might suggest. I hope so anyway.
Max Drei-09.04.2026
I have never been in favour of deleting parts of the description at grant or before maintenance in amended form, hence my suggestion has always been to mark accordingly embodiments not falling under the definition of the matter for which protection is sought derivable from the claim.
If these passages describe subject matter that is plainly equivalent to what is literally claimed, then those parts do not have to be marked accordingly as such, but some information is necessary for third parties.
It would say that a good drafter would either draft a claim encompassing all the various embodiments, for instance of the “means + function type”, or file independent claims for each embodiment, provided unity of invention is given.
If unity of invention is not given before grant, then the only solution before grant, is to file a series of divisional applications. If unity of invention is given, and the applicant is not willing to claim those equivalents as such, then a STIN mention ought to be introduced on the front page of the patent specification, making clear that there are equivalent embodiments in the description.
As unity of invention is not a problem in opposition after amendment, the proprietor is allowed to file independent claims, each covering those equivalent embodiments, whether unity is given or not. In opposition, whether unity of invention is given or not, and the proprietor is not willing to claim those equivalents as such, then a STIN mention ought to be introduced on the front page of the new patent specification, making clear that there are equivalent embodiments in the description.
The aim is simply to provide the necessary clarity for third parties, so that they do not have to dig deep into the description to determine what is actually defined as matter for which protection is thought.
When, after examination or opposition, embodiments are not equivalent to those falling under the definition of the matter for which protection is thought, and are clearly not any longer patentable, then a corresponding acknowledgement is a must.
Leaving equivalents in the description and not claiming them, is in my opinion, a rare occurrence. It cannot justify an absolute ban on the adaptation of the description.
@ BF-09.04.2026.
I read with great attention Mr Snodin’s TPO. At least is less aggressive and derogatory as the anonymous TPO. His conclusions are however aligned with those of the anonymous TPO.
What is interesting to note is that Mr Snodin is recycling what he has said in its amicus curiae for G 1/24: “The purpose of using the description to interpret the claims is to ensure that the meaning afforded to the terms used in the claims takes account of context provided by the description. This is in accordance with the general principle of law that a document must be read as a whole”.
He also considers that support in only a one way street: what is in the claim should be somewhere in the description. That the description contains more is for the benefit of claim interpretation as anything in the description can be read in the claim. This is also the possibility envisaged in T 439/22, but not in all other board’s decisions applying G 1/24. What discussions at the PCT level imply for the EPO remains to be seen.
What is at stake in G 1/25 is not merely interpretation of the claims but relates to a proper and reliable definition of the matter for which protection is thought. This is different to the interpretation of the claims under Art 69. A claim can be perfectly clear as such, but the description may give another definition of the matter for which protection is sought. This is the problem which should be solved with G 1/25.
It cannot be that the description allows a different definition of the matter for which protection is thought than that of the claim or vice versa.
When applying G 1/24 as all bords have done, but in T 439/22, there is a first alignment between claims and description: if the applicant/proprietor wants the definition of the matter for which protection is thought present in the description, then the claim has to be amended accordingly. It could also go the other way round.
Once the description and the claim contain a priori the same definition of the matter for which protection is thought, then what has to be checked is whether other discrepancies allowing a different definition of the matter for which protection is thought.
This is why I do not agree, neither with Mr Snodin, nor with the anonymous TPO and not with T 56/21, Reasons 53 onwards. It might be correct to claim that Art 84 states a requirement in respect of the claims and not of the description, but in combination with R 42(1,c) it is not possible to allege that support is only the one way street going from the claims to the description. T 56/21 went one step to far to say that when the claim is clear, no need to look at the description. This is a fundamental flaw in the reasoning in T 56/2 in view of G 1/24.
You cannot tell me that a feature contained in the original independent claim but deleted due to prior art, can be left in the description to still represent the matter for which protection is thought. By doing this you thereby ignore the description, and this is what T 56/21 claims and is therfore not consistent with G 1/24. The same goes for an optional feature in the description used to limit an independent claim. Once it has limited the independent claim, ithe same feature cannot remain as being optional in the description.
I invite you to bring forward a compelling argumentation against these views.
Let us enquire a little further into the meaning of “the protection sought” in the context of, first, the A publication and then, second, the B publication.
In the A publication we see a claim, supported by an enabling disclosure. There might be an “inconsistency” between the claim and the description, but more likely not. Now assume a prosecution at the EPO where the claim gets amended, to something narrower. That narrows the definition of the protection sought. But if the description stays unamended, is will very likely be an “inconsistency” between the narrowed claim and the original description.
But is that enough, to destroy legal certainty, or offend one or other provision of the EPC?
Yes, I say, if the description defines a technical feature in a way that contradicts what the skilled reader derives, upon reading the claim. Treating the B publication as a single document, the reader cannot derive a clear and unambiguous meaning to the claim.
But what if the reader of the B publication understands and takes into consideration that the protection sought has been narrowed down during prosecution, from what the A publication declared to what the granted claim defines. Might that reader be able to derive a clear meaning for the claim, even when the description of the B publication still continues to memorialise what Applicant thought was “the invention” at the time of filing the application.
And in the courts, might it not help rather than hinder reflections on Art 69 EPC, to know from the patent publication what protection was originally sought and what the EPO cut it down to.
I’m looking for a practical way for inventors reliably to get from the EPO a scope of protection commensurate with their contribution to the art, after the EPO has cut down the scope of protection during prosecution, without the inventor instead disappearing, either prior to grant within the notorious Bermuda Triangle formed by Art 54, 84 and 123(2) EPC, or after grant within the inescapable 123 2/3 trap. EBA please try to find a solution that strikes the right balance between inventor and public, and recognises the complexity of ameding at the EPO a patent application after its filing date..
Max Drei,
I agree with you that, if the claim has been narrowed, any broader definition in the description of the matter for which protection has been sought, represents a clear cut inconsistency.
Where I cannot follow you, is when you say in the same breadth that, when looking at the B publication as a single document, the reader cannot derive a clear and unambiguous meaning to the claim. In other words, you consider that the granted claim is unclear, but limited.
First, in this situation a patent should not have been granted, but this can happen. In post grant or post opposition proceedings, the court deciding upon infringement or validity will have to interpret the claim in the light of the description.
This is also done at the EPO when the original claim is unclear. Merely coming up with Art 84 is rarely bringing matters forward. It is more efficient to interpret the claim and, based on this interpretation, object N or IS.
What you are apparently aiming at is to introduce file history in the appreciation of the terms of the claims. A subsequently acting court is perfectly entitled to do so, and this is beyond the reach of the EPO.
At the EPO’s level, file history is not a way allowed to define what the matter for which protection is sought. I repeated it many times, the day a kind of file wrapper estoppel is introduced in the EPC, then no adaptation of the description will be required.
I maintain that, if one or more features/embodiments contained in the original independent claim were deleted due to prior art, it cannot be left in the description that those features/embodiments still represent the matter for which protection is thought.
For the same token, when an optional feature in the original description is used to limit an independent claim, then, the same feature cannot remain as being optional in the description.
I do honestly not see where is the problem. By aligning the description with the claims, or vice-versa, I fail to see that the applicant/proprietor will be bereaved of its contribution to the art. If the claim is clear at grant or at maintenance after opposition, a subsequent judicial authority will be in a position to determine what the scope of protection could well be.
The EPO has just to grant or maintain a patent in which the matter for which protection is thought is properly defined. This is necessary so that the applicant/proprietor cannot later switch from one definition of the matter for which protection is sought to another one. This is the combined effect of Art 84 and R 42(1,c).
That in case of litigation the party accused of infringement or an opponent will look three times at the original disclosure and at the granted patent, and this front to back and back to front, is not something the EPO has to be concerned with.
Daniel once more our dialogue founders on a difference of understanding of the meaning of one word, here “definition”. Never mind though. I have written what I wanted to say and will leave it at that, hoping that other readers will appreciate what I have tried to communicate. Whatever the EBA decides, we shall all have to work with it and that will be OK.
Max Drei,
I am understanding your position, but I cannot share it.
For the readership, both your point of view, and that of other members of the profession, as well as mine, are on the table. We will now have to wait until the EBA has decided. Whether we like it or not, we will indeed have to work with the order of the EBA.
I do however have no understanding for people wanting, like for instance M. Snodin, Anon Y. Mouse, and the writer of the anonymous TPO, to tell the EBA what it has to do and how it has to interpret the EPC. To say the least this is preposterous.
I am aware that my position is a minority position amongst the profession. It does however turn out that nothing what I have heard up to now against adaptation of the description is, in my eyes, compelling. Adaptation of the description, corresponds to what I have been taught and what I gave down the line to the examiners under my responsibility.
You have tried to argue about the deeper meaning of “definition (, and of “clarity”). With a mind willing to understand, I do not see what it could bring. I have given two clear and simple examples in which, against all odds, I cannot see what could be dangerous for the applicant/proprietor if the description is amended.
That representatives express some fears, is understandable, as some divisions of first instance have had a rather academic approach f the matter, but with a bit of common sense and good will, I do honestly not see where a problem could creep up.
If the applicant/proprietor does present an amended description and the deciding body has finds and has explained why the amendment is not in conformity with the EPC, then the deciding body has no other alternative than to refuse the application or to revoke the patent.
In my professional career I have been confronted with proprietors or opponents who only wanted an adaptation of the description or an acknowledgement of the prior in order to have a better position in future litigation. Needless to say that they were not successful.
When I now see the outcry from the profession, some being violently against adaptation of the description, I cannot refrain a snigger. The more the profession is against adaptation of the description, the more I am led to conclude that what the profession wants is to leave everything in the description as fuzzy as possible so as to try to get more out, should it come to litigation.
This is not something what the EPO should tolerate. Rather, the duty of the EPO is to bring a specification in a shape in which the proprietor cannot eat its cake and have it. It is as simple as that.
I had decided not to write anything more, Daniel. But, for what it’s worth, I will try to explain what I see as two meanings of the word “definition”.
In imaging technology, we see more and more references to “definition”. HD, UHD etc etc. In the context of dictionaries, however, a “definition” tells us what a word means.
In the context of the “description” of a B publication, the word “definition” might refer to a dictionary definition., in the description, of a word found in claim 1. But alternatively, it could mean that the description as a whole adds colour, depth and fine detail to what is “the subject matter for which protection is sought”, what the inventor thought was their invention when filing the patent application.
I agree with you that if the description contains a dictionary definition of a word found also in the claim, any resulting contradiction between the claim and the description is objectionable. Conversely, passages in the description of the B publication that provide detail depth to the disclosure of the invention for me fall short of a contradiction that offends a provision of the EPC. To go over old ground, (my “third way”), one can have a description which includes detailed decription of specific apparatus or methods, without saying that is definitively is, or is not, within the scope of protection. I resist insistence, as a condition of grant, that I say definitvely, in a binary way, whether each drawing Figure is inside or outside the scope of protection.
2026.04.10-Comment from Anon Y. Mouse – Deleted by error.
Here is the restored comment:
DX Thomas raises his familiar concerns.
He focuses on the question of how to define “the matter for which protection is sought”. This is in Article 84, sentence 1, which assigns the definitional role exclusively to the claims.
The description does not co-define the matter for which protection is sought. It supports the claims that do define that subject matter. It seems to me that DX Thomas’ entire framework depends on treating the description as having equivalent definitional status to the claims, which is precisely what the statutory text does not say.
Rule 42(1)(c) does not help his case. It requires the description to disclose “the invention, as claimed” to a sufficient standard. This governs the standard of disclosure, rather than establishing some sort of greater role for the description in defining the subject matter for which protection is sought.
G 1/24 itself confirms at point 8 of the reasons that “Article 84 EPC addresses the content of the patent application and is formal in nature, does not mention the invention and provides no guidance on how to interpret claims.” Yes, G 1/24 requires the claims to be interpreted in consultation with the description, but it does not overrule the fact that Article 84 assigns the claims – and only the claims – the role of defining the subject matter for which protection is sought.
DX Thomas cites his favourite example of a feature deleted from the claim due to prior art left in the description. He says this cannot be permitted because it would still represent the matter for which protection is sought. But a skilled person reading claims and description together in order to understand the claim (as G 1/24 requires) will immediately understand that a feature present in the description but absent from the claim is not part of the claimed subject matter. Courts handle this routinely; see for example the BGH Okklusionsvorrichtung (X ZR 16/09) case, which establishes as a principle of claim interpretation that parts of the description not reflected in the claims are generally not within the scope of protection. More seriously, if deletion of that description passage is performed, T 439/22 demonstrates that it may violate Article 123(3) by removing an implicit restriction on the claim. The mandatory adaptation that DX Thomas advocates can therefore produce the very legal violation it is meant to prevent.
It is also worth considering the Travaux Préparatoires of the EPC as cited in Rockwool’s latest submissions in connection with G 1/25. The passages cited there confirm that the support requirement was intended solely to ensure claims remain within the bounds of the original disclosure, with no mention of any requirement to align the description to the claims.
The above is offered in the spirit of providing the compelling counter-narrative that has been requested. However, none of these points are new. They have been made abundantly and in more detail by other parties.
@ Anon Y. Mouse
Your last comment has been deleted by error.
As far s the EPO is concerned, it is irrelevant that in the BGH Okklusionsvorrichtung (X ZR 16/09) case, establishes as a principle of claim interpretation that parts of the description not reflected in the claims are generally not within the scope of protection. They might not be within the scope of protection, but they still attempt to define the matter for which protection is sought.
The way post grant or post opposition judicial bodies interpret the claims is their good right and does not free the EPO to insure that what is in the claim has to be supported by the description. If the description comprises statements allowing to broaden or limit the definition of the matter for which protection is thought, then there is an inconsistency between claims and description.
In the famous Agfa/Gucci case, this was exactly the position taken by the UPC-CFI LD Hamburg. The court was not fooled by the proprietor having left in the description chromatic, whilst the claim had been limited to achromatic, see Headnote 3. The LD Hamburg noted that there is an inconsistency between the description and the independent claim. The LD Hamburg nevertheless added that “There are several passages mentioning “chromatic” in the specification of the patent that should have been deleted”.
In decision UPC_CFI_363/2024 of 01.08.2025, the LD Paris in § 1 of the abstract, the court further noted that the addition of the words “placed edge to edge” in claim 1 should have led, at the time of the grant of the said patent, the removal of the so-called “stacked” embodiment of the invention mentioned in the description, as this embodiment does not fall within the scope of protection of the patent’s teachings as ultimately set out in the claims. The LD Paris referred to the Agfa/Gucci case.
It is difficult to see how deleting chromatic from the description would end up in a infringement of Art 123(2) or 123(3). I do need some compelling argumentation and not a mere reference to T 439/22.
It would be interesting to see how the board in T 439/22 would deal in opposition with a claim limited to achromatic, where in the description chromatic would have been left. When following the reasoning in T 439/22 this would mean that the claim would cover at the same time chromatic (broader definition of the matter for which protection is sought), as well as achromatic (limited definition of the matter for which protection is sought). In catchword 3, the board specified that the skilled person will read terms in the claim in the sense of the definition, taking into account both the broadening and limiting aspects. Welcome to a nice mess!
This shows best that T 439/22 cannot be followed, but all the other decisions applying G 1/24 have shown the correct way to follow. It is for the applicant/proprietor to decide which definition of the matter for which protection is sought, but he cannot leave different definitions in the claims and in the description. In this respect, I maintain that G 1/25 can be considered the logical follow-up of G 1/24.
As far as Art 84 is concerned, the travaux préparatoires do not give much away, other than Art 6 PCT has been amended by deletion of “fully” when it became Art 84 EPC.
There is thus absolutely nothing compelling in your comment, contrary to what you might think. To me your alleged “compelling counter-narrative” reminds me of a deflated balloon.
When mentioning case law, you should come with more up-to-date case law and not a decision of the BGH of 07.06.2011. This decision has been superseded with more recent case law, for instance from the CFI-UPC.
I am inclined to consider that the topic is by now exhausted, and it appears wise to wait for the OP and then for the decision.
Unless really ground braking comments have to be uttered, it appears best to stop the discussion here.
What had to be said has been said and the different opinions are crystal clear.
@ Max Drei-12.04.2026
The problem is that you are using the word definition in a rather specific way. UD, UHD, HF, VHF are all terms having a precise meaning, like what can be found in dictionaries and standard books.
Defining the matter for which protection is sought, is something different. It defines a set of features, whereby each might have a definition in a dictionary, but which as a whole, provide a technical effect which was not obtained beforehand or obtained with other means. This set of features has at least to be new but also inventive.
If a claimed feature is replaced in the description by a statement giving the claimed feature a different definition, be it broadening or limiting, then an objection has to be raised as the definition of the matter for which protection is sought is not properly defined. If this definition is different in the description and in the claims. this is clearly objectionable.
It is not necessarily a contradiction, it suffices for it to give rise to doubts about what is the matter for which protection is sought.
Passages in the description of the B publication that provide detail depth to the disclosure of the invention fall indeed short of a contradiction that offends a provision of the EPC. If the description includes a detailed description of specific apparatus or methods, or whistles and bells which could be added, those should then be subject-matter of dependent claims, provided they fall under the broad definition of an allowable claim.
If out of three originally disclosed embodiments, two are anticipated by prior art, it is not acceptable to maintain in the description that those embodiments still represent matter for which protection is thought. This would defy any logic and would render claims superfluous. .
Anything else in the description which cannot be transformed in dependent claims should be marked accordingly.
You might not be willing to say that it definitively is, or is not, within the definition of the matter for which protection is thought. I note first that, speaking here about the scope of protection is totally misplaced. You might also resist insistence, as a condition of grant, but then your client will have to bear the consequences, i.e. refusal of the application.
The scope of protection is something different and it should by now be clear, that the scope of protection is not for the EPO to decide, but in applying Art 123(3). Removing a feature from a granted claim invariably increases the scope of protection, whatever it might be, in that the breadth of a claim is in inverse proportion of the number of its features, hence the inescapable trap.
Just think of Epilady/Improver. The rotating rubber rod with slits fell within the scope of protection of the rotating spring as subsequent jurisdictions (but the UK ones) found out, but it was neither disclosed nor claimed. Should this embodiment have been suggested in the original description, he could have stayed in, either as dependent claim of a means + function claim or as a further independent claim. Am applicant would have been entitled to resist deletion and any reasonable examiner would have accepted that it should not be deleted.
This is why I find G 1/24 so important. It should induce a return back to basics for EDs. If the EPO wants to deliver high quality patents, as claimed by the upper management, then production might have to decreases slightly, but I am not even sure of it.
G 1/24 actually forces EDs to consult the description before search and examination, and before grant.
Before search and examination it will allow the SD and the ED to realise the full breadth of the original disclosure and act accordingly when searching and examining. Before grant it will allow to free the description of ballast which has nothing to do with the definition of the matter for which protection is sought. Depending on the situation this ballast has to be deleted or marked accordingly.
Independently of G 1/24, the SD and the ED should have spotted that the description contained a broader definition of “gathered sheet” and searched accordingly. Before grant, they should have spotted that the description contained a broader definition of “gathered sheet” and should have queried the difference. With G 1/24, it has become a duty of the SD and ED to consult the description.