Following the release of G1/25, questions have been circling as to what the decision means for the pending referral in G 1/26 and what the consequences (good or bad) are for applicants and patentees. On G1/26, some of the questions have been answered, but arguably only because the EBA considers these questions already answered by G1/24. With respect to the consequences of G1/25 per se, this Kat finds herself (not for the first time...) out of step with the initial consensus. However, a closer reading of G1/25 has only strengthened this Kat’s view that G1/25 is not to be welcomed by applicants or patentees.
Legal background: Claim interpretation and description amendments
G1/24 held that the description and drawings shall always be consulted to interpret the claims (IPKat). Following G1/24, the Boards of Appeal were left with the task of interpreting what “consulting” the description actually entails. Various approaches have so far emerged, including consult but never broaden and/or narrow (IPKat), or a “holistic” approach permitting both broadening and narrowing (IPKat).
There were subsequently two new referrals, G1/25 (description amendments) and G1/26 (claim interpretation and added matter). Given what G1/25 says about G1/24 (IPKat), it arguably answers some of the questions in G1/26.
G1/26: Claim interpretation and added matter
In T 0873/24, the referring decision in G 1/26, the granted claim in question specified a titanium to nitrogen ratio “in excess of 3.42” without saying what kind of ratio it meant. Read on the claim alone, the term covered every ratio making technical sense to the skilled reader. This interpretation was considered by the Board of Appeal to add matter. The description on the other hand gave the figure by weight, potentially removing the added matter issue.
The Board of Appeal therefore thought it pertinent to ask the EBA for some clarification and referred three questions. There has been some debate as to whether G1/25 has now already answered one or more of these questions.
G1/26 Question 1: Admissibility
Question 1 of G1/26 is an admissibility question, asking when an EBA decision may be considered “required” in order to ensure uniform application of the law, or if a point of law of fundamental importance, and thereby justify a referral. G 1/25 does not provide a view.
G1/26 Question 2: How should the description be used to interpret the claims
Question 2(a) asks whether starting with the claims as the basis for assessing patentabiltiy, "precludes a feature which is only disclosed in the description or the drawings of a patent from being read into the meaning of a granted claim, in particular if this leads to a restrictive reading of terms used in the claim". This question is answered to some degree in G1/25. R.10 holds that the skilled person “will try to take a definition found in the description at face value”, reading the claim term in that sense “taking into account both the broadening and limiting aspects” (r. 10, emphasis added). Question 2(a) is therefore answered in the negative. The primacy of the claims does not rule out narrowing features from the description being read into the meaning of the claim in general.
The Board of Appeal approach in T 439/22 (originally an outlying decision) is therefore supported by the EBA (IPKat). Interestingly, this case is also the case that found that description amendments can extend protection from an added matter standpoint.
![]() |
| Consequential dominos? |
Question 2(b) first asks whether reading the claims and description is a unitary process, or whether the claims are read first. This first limb of Question 2(b) is clearly answered in G 1/25. According to the EBA there are no "different stages of interpretation" (r.9).
Question 2(b) of G1/26 then asks whether the claim being the starting point and the basis for assessing patentability rules out only those interpretations which can be derived from the patent as a whole but would clearly contradict the general technical understanding of the terms used in the claim. On this point, the EBA found that description and drawings “cannot be used to impose on the claim a limitation or expansion for which the claim wording provides no basis” (r. 10, emphasis added). This would seem to imply that the description cannot be used to change the meaning of terms used in the claims to something other than their normal definition. We are, however, now left wondering what “no basis” means. In particular, is “no basis” meant merely to exclude new or unusual definitions (unusual parameters), or to exclude any departure from the normal meaning of the words used in the claims? Or does “no basis” mean that the interpretation must be clear and unambiguous to a skilled person from the claim language?
For the referring case in G1/26, the question thus becomes whether “in excess of 3.42” provides a basis in the claim wording for reading in “by weight”. On one view a bare ratio is meaningless without units, so the claim wording itself calls for units to be supplied and the description does no more than supply them. By contrast, the view could be taken that “by weight” is a limitation for which the claim does not provide clear and unambiguous basis, that r. 10 of G1/25 forbids reading such a limitation into the claim, and that the claim as it stands adds matter.
G1/26: The added matter questions
Question 3(a) of G 1/26 asks whether, when assessing compliance with Article 123(2) EPC, a claim term must be tested against every interpretation that makes technical sense to the skilled reader on the basis of the claim alone. Question 3(b) asks, if it need not, whether it is enough that only those interpretations arrived at against the background of the specification as a whole are directly and unambiguously derivable from the application as filed. In this Kat’s view, these questions are unanswered, or at the very least not yet clearly answered by G1/25. Interestingly, these questions recall the guidance sent to Examiner's on application of G1/24, which requires searches to be carried out on all the possible interpretations of the claim, implying that the EPO's current view is that alternative interpretations must be taken into account in some respect (IPKat).
Why G1/25 is bad for applicants and patentees
Since G 1/25 was issued last week, there has been much debate over whether G1//25 will make the description adaptation requirement worse or better for applicants and patentees. This Kat is of the view that it has made things worse (IPKat).
First, the risk that a description amendment is itself the invalidity ground has been reinforced in view of Article 123(2) EPC and Article 123(3) EPC. In T 439/22 the deletion of a definition of “gathered” from the description was held to extend the protection conferred, leaving the patentee in an inescapable added matter trap (IPKat). G1/25 establishes that the description can be used to narrow claim language. If you delete a narrowing definition, even at the invitation of the Examining Division, G1/25 confirms that the deletion is capable of broadening the granted claim. It also arguably makes that outcome more likely. This Kat is therefore at a loss as to why anyone would voluntarily make amendments to the description in view of this risk, reasoned or otherwise (unless at a last ditch attempt at trying to cover a competitor product with amendment by the back-door, see Ensygnia v Shell ([2023] EWHC 1495 (Pat)), IPKat).
Description amendments are also now more consequential post-grant. Any amendment will be seized on by opponents as altering claim interpretation, and patentees will therefore need to be prepared to defend each one from an added matter attack.
We can also expect that description amendments will be brought up in litigation in the form of file wrapper estoppel. The UPC Court of Appeal in VusionGroup v Hanshow notably declined to rule out prosecution history as an aid to construction. Description amendments will therefore likely be cited as influential on intended claim scope.
G1/25 may also have an impact even beyond Europe. In the US, claim interpretation takes account of the description and the prosecution history, including what amendments were made and what arguments were submitted, and the file wrapper does not stop at the US border. In K-fee v Nespresso (Case No. 22-2042, Fed. Cir., 26 December 2023) the Federal Circuit treated statements the Patentee had made about the meaning of “barcode” in EPO opposition proceedings as properly part of the material for construing the corresponding US claims (IPKat).
Therefore, if you care about IP strategy beyond the formalities of examination and if you care about more than just Europe, G1/25 is not good for applicants and patentees. G1/25 may, reduce the formalities admin burden of description amendments at the Rule 71(3) stage, but at the price that any amendments that are made, or even not made, may now be given more weight with respect to claim interpretation post-grant.
What does G 1/25 actually mean for description amendments?
There also remains some debate as to whether G1/25 will actually make adaptation of the description a less onerous requirement for applicants in view of Article 84 EPC.
G1/25 confirms that Article 84 EPC can be used as the basis for an adaptation of the description. Article 84 EPC states that “The claims shall define the matter for which protection is sought. They shall be clear and concise and be supported by the description.”
Established case law to date has been that the claims must be clear in themselves for a person skilled in the art with common general knowledge of the technical field in question, without the need to refer to the description. However, G1/25 does not follow this path. The decisions taking that view, the EBA says, “proceed on the basis that Article 84 EPC concerns only the claims and that clarity must be assessed without reference to the description”, and to the extent that premise excludes the interpretative role of the description recognised in G 1/24, “it cannot be followed” (r. 29). The principles of G1/24 instead “apply where the assessment under Article 84 EPC requires the meaning of the claims to be determined” (r. 33).
As such, G1/25 may rescue applicants caught between added matter and clarity, because a definition sitting in the description can now be used to make a claim term clear (so “catastrophic comma loss” is less of a risk). On the other hand, the description can now directly be used by the Examining Division as the basis of a clarity objection.
R.20 of G 1/25, which many commentators have pointed to as supporting a more lenient description amendment requirement, states:
“An inconsistency is not established merely because the description, including any drawings, contains a technical teaching, examples, or embodiments that do not fall within the claimed subject-matter.” (r. 20)
So far so good. However, the same paragraph then goes on to say:
“If, on the other hand, due to such an incompatibility, it is unclear whether a technical teaching, examples, or embodiments do, or do not fall within the claimed subject matter, an inconsistency exists.” (r. 20). R. 35 then provides that where it is “unclear whether information, examples, subject-matter or embodiments are or are not within the scope of the claim, then it cannot be said that the claim is supported by the description” (r. 35), i.e. they offend Article 84 EPC.
Importantly, r. 35 cannot be read as a free-standing licence to demand deletions. It opens with the words “applying the above-defined concept of inconsistency”, which places it downstream of r. 18 to 20 and means that it presupposes an inconsistency already established. However, r. 20 and r. 35 both ask the same question of whether it is unclear that matter falls inside or outside the claim, once to establish the inconsistency and once to decide on Article 84 EPC. As such, the two-step test that commentators have taken from G1/25, inconsistency first and non-compliance with an EPC provision second, therefore can be said to collapse into a single step whenever the provision relied on is Article 84 EPC.
Final thoughts
For this Kat, at the heart of G1/25 is the question of not only how we think a skilled person understands terms used in the art, but how we think a skilled person reads and understands a patent document. After all, all of the discussion presupposes that a skilled person can navigate and understand the distinction between the claims and description, which are themselves artificial creations of patent law. A skilled person reading a patent would also see the abstract, and as a scientist used to reading journal articles, may get very confused by this. However, there is no legal basis for the patent abstract leading to a lack of clarity or thereby needing amendment.
Following G1/25, it seems that everything on claim interpretation now depends on what counts as “basis” for a limitation or expansion. Mandatory description adaptation now arguably depends, inter alia, on when it is “unclear” whether an embodiment falls within the scope of the claim. However, neither term is defined and we wait to see what the Boards of Appeal, and Examiners, will do. In the meantime, applicants and patentees should take extra care with any description amendments and any arguments made in relation thereto. Any position taken for or against description amendments is now potential material on the file that an opponent or a defendant will be more than happy to cite.
- Don’t shoot yourself in the foot: European file history in US patent claim interpretation (K-fee v Nespresso) (January 2024)
- No back-pedaling on prosecution disclaimers (Azurity v. Alkem, Fed. Cir. Case No. 23-1977) (April 2025)
- EBA decides G1/24 on claim interpretation: The description should always be consulted (June 2025)
- BREAKING: Referral on description amendments finally confirmed! (G1/25 - “Hydroponics”) (July 2025)
- Divergence between the UPC and EPO on claim interpretation and description definitions (Agfa v Gucci, UPC_CFI_278/2023) (August 2025)
- Description amendments can extend protection (T 439/22) (February 2026)
- New referral on claim interpretation and relevance of G1/24 to added matter confirmed (G1/26) (June 2026)
- Claim construction checkpoints: Application as filed versus granted patent (T 0715/24) (August 2026)
- BREAKING: When real doubt hits - G1/25 changes the description adaptation requirement for the worse (September 2026)
Reviewed by Dr Rose Hughes
on
Tuesday, September 08, 2026
Rating:

OK. Germany's Supreme Court follows legal scholars in the USA to stress that:
ReplyDelete„Der Bundesgerichtshof hat die Aufgabe, die Rechtsprechung zu vereinheitlichen und das Recht fortzubilden.“
That is: The task of the supreme court is to reduce complexity and bring instead the simplicity and clarity that enables an orderly progress and further development of the jurisprudence.
In G1/25, you may well think as I do, that in this social task, the EBA has obviously, patently, disappointingly failed. Somewhat ironic, don't you think, in a decision addressing, of all things, Art 84, EPC.
What alarms me is the notion that the "holistic approach" allows EPO Examiners to contend that:
1) the claim under examination, construed holistically, has a level of "clarity" that fails to reach 100%
2) therefore, there exists an "inconsistency" between that claim and the dcescription
3) which inconsistency is the cause of "real doubt" about what the claim "covers"
4) whereby only an amendment can save the patent application from refusal.
One can see G1/25 as a decision drafted by an EBA stacked with judges who, in their respective "day jobs" must decide, downstream, what a claim covers and what it does not. But who on the EBA was capable of explaining to their colleagues how the world looks to patent applicants. Perhaps nobody on the EBA sees any need to simplify the world for patent applicants. Rather, all their efforts are to simplify the world for "the public" that is, those threatened with patent infringement. The corporate applicant community has immense lobby power. How much does "the public" have? Perhaps "the public" needs all the help it can get from the EBA?
Max, please forgive the direct and slightly impolite question, but what do you actually believe?
DeleteI ask because both here and on another blog your comments seem to blow this way and that, flattering DXT on occasion and gently pushing back on others; suggesting that G 1/25 is an elegant compromise in one comment, flagging possible problems in another and so on. And now on this particular blogpost you cast a very negative light on G 1/25 despite you comments on the previous post on this very blog suggesting a warmer welcome.
Has your opinion changed as you've meditated on the decision over the last few days? Are you simply posting different takes to elicit different reactions from different quarters? As you've told us, you're retired so no longer have any stake in the outcome, so what is the end game here?
You've long played a very good game of teasing out positions and contradictions from either side, but there comes a point at which the two sides' opinions are sufficiently clear that playing the go-between doesn't yield any further useful information (see, for instance, Daniel deleting comments he disagrees with and declaring his interlocutors "stupid" or unfit to serve in the profession - remarks which I think are to be deplored and which I'm astonished have gone without reprimand from the hosts of his blog, but that's by the by).
I ask this as someone who respects your experience and long history of thoughtful contributions to the debates on many and various topics over at least the nearly two decades I've been in practice, if not longer. But just for once, now that the EBA has spoken and the battle lines have been clarified, I'm curious to know which side of the fence you would come down on, on this topic at the very least.
Thanks for asking, AYM. Over the years, I have found it necessary from time to time, to change my opinion. Here, I can see both sides of the argument and find myself unable yet to jump down from the fence on one side or the other. Judges have to jump, every time they try a case. It is, or ought to be, in difficult cases a devilish hard thing to decide on which side to throw themselves down. I lack the self-confidence to be a judge.
DeleteInstead, as you have noted, I seek to provoke answers to my questions, that will reveal with greater clarity the difficulties each side of the fence. Sometimes my provocations are seen as snarky, or otherwise upsetting. I don't mean them to be.
If I were given the job of deciding which of the "battle lines" should win, I confess that I would still not know how to decide.
I used to think that binding precedent is the best way to deliver legal certainty. But in the law of patent validity, it seems to me that the EPO's Darwinian evolution of their case law has delivered better results, in the long run. So too here, I suspect.
@ Max Drei-08.09.2026 at 18.38.00 GMT+1
DeleteIt does not appear correct to allege that in G 1/25, “the EBA has obviously, patently, disappointingly failed”. That it dealt with Art 84 is, to a large extent, due to T 56/21 and the like, and to the fact that most of the amicus curiae briefs refusing the adaptation of the description mainly based their argumentation on the premise that Art 84 is only valid for the claims and has no relation with the description.
It should therefore not come as surprise that Art 84 applies in both ways. This is what I have been always taught and conveyed to the examiners I was in charge.
The “holistic approach” you complain about, will, in my opinion, not bring all the disaster on applicants/proprietors you foresee.
The holistic approach, as now clarified in G 1/25, means that a broadening or limiting interpretation of all embodiments originally disclosed in the description will have to replace in the claim, the definition of the matter for which protection is sought in the original description or vice-versa. This is at the choice of the applicant/proprietor.
In any case there cannot be in the same patent two different definitions in the claim and in the description of the matter for which protection is sought. Those times have gone with G 1/25.
I allow myself to go as far as to say that a correct application of G 1/24 will automatically resolve a great deal of inconsistencies. However, the boards will never “read” in the claims, limitations only found in specific embodiments of the invention or the disclosure. This is by now well settled in the case law, cf. T 2027/23 and the like.
The corporate applicant community has indeed an immense lobby power, but this power did not convince the EBA that the description does not need to be amended once the claim is limited due to prior art.
I have said it many times, the day Art 84 is amended, and at the same time a kind of file wrapper estoppel will be entered in the EPC, then the description will not need adaptation.
It is also not correct to claim that the EBA stacked with judges who do not care about how the world looks to patent applicants. This is deeply unfair bordering on insult.
The EPO at large, and the boards especially are certainly aware that a patent is not a piece of paper to be hung in a lounge or the entrance hall of a company, but represents an economic asset.
The EPO has been set up to deliver solid patents, i.e. patents for which the proprietor cannot be left the choice of the interpretation which suits it best depending on the circumstances.
A claim can be perfectly clear as such, but there can well be an inconsistency between the claim and the description, if the supporting description allows to give an interpretation which does not tie up with the definition of the matter for which protection is sought in the claim. Is this so difficult to understand?
I have said it for a long time, the EBA should be a jurisdiction totally independent of the “normal” boards of appeal. This does not mean that the EBA should exclusively comprise LQM and TQM from outside, but in my opinion, membership of the boards and membership of the EBA within the EPO should be distinct. The legislator, for whatever reason, has not decided this way and we have to accept it. This is however not a reason to speak so badly about the members of the EBA.
I generally agree with the analysis in the post. What I do not understand, however, is why Dr Hughes and most commentators focus almost entirely on r. 20 and r. 35 and very little attention is given to r. 38. In my view, that part of the decision could be used to justify the vast majority of description amendments requested during prosecution...
ReplyDeletePassing By: the problem with r. 38 is that it only bites if the relevant "statement" in the description "conflicts with the fulfilment of the requirement of non-obviousness". This therefore demands reasoning which demonstrates why the presence of that "statement" in the description leads to the claim not meeting the requirements of Article 56 EPC. Frankly, I do not see why a disclosure relating to a different embodiment (ie lacking a limiting feature found in the claims) would lead to non-compliance of the claims with Article 56 EPC. Can you?
DeleteIn any event, all of this turns upon the existence of an "inconsistency". For reasons that I explained in a comment on 8 September on Rose's other recent G 1/25 post, it is difficult (if not impossible) to make sense of the EBA's definition of "inconsistency". Do you have any insights which you can share and that may help me (and, perhaps, others) to make sense of what the EBA is trying to convey in r. 18?
What was the Enlarged Board trying to convey in r. 18? Perhaps a public policy objective, that under the EPC, where a lack of clarity is not a basis for attacking the validity of an already granted patent, it is the duty of a Patent Office, vigorously and rigorously, to ensure that no claim ever gets to issue that is in any way less than, directly and unambiguously, to a level of 100%, clear in meaning, and that meaning not so much for lawyers as for the vast numbers of European engineers and scientists who will be reading the patent and asking themselves the question of over-riding importance, namely, what infringes and what does not.
DeleteEverything else is of a lesser degree of importance, the Board is telling us. Europe shall lead the world, henceforth, in distinguishing claims with merit from claims with none, swiftly enforcing the former, and swiftly extinguishing the latter. Good, eh?
Dear Proof, r. 38 states "and the statement thus conflicts with the fulfilment of the requirement of non-obviousness". In view of the "thus" I did not see the statement as imposing an additional condition that must be met but rather as elaborating on what comes before, "a statement [...] expresses a technical teaching reflecting the claim before the amendment". I accept your view that the decision might be read differently, we'll have to see how it will be applied by the Boards. On the question of what "inconsistency " precisely means I suspect a lot of ink will be spilled in the future, unfortunately I am not in the position of providing any additional insight.
DeleteSure. Because the EPO is so omniscient and perfect with its examination of applications that it always spots a lack of clarity that might elude us lesser mortals unless and until we become aware of crucial prior art and/or expert evidence that does not form part of the examination file.
DeleteIn any event, the EBA is not, and should never be, in the position to establish public policy. If, on the other hand, the EBA believes that a public policy of insisting upon an unachievable ideal (of 100% clarity) is derivable from the provisions of the EPC, then they must be looking at a version of the EPC that is not available to the rest of us!
Passing By: it may not be clear to you from "thus" in r. 38, but what about the explanation which is provided in r. 39? Does that not make the matter crystal-clear?
DeleteDear Proof, to be honest I find r.39 anything but crystal clear. The "so interpreted" likely makes reference to G 1/24, which according to G 1/25 requires to consider description and claims together for claim interpretation. Then, I am not really sure when the infamous inconsistency "materially affects" the interpretative exercise, since r.18 already defines inconsistency by reference to that very same exercise, and even then the claim only "may" fail to meet the patentability requirements.
DeleteI would like to emphasise that I am not trying to convince you that your reading of the decision is wrong or that my initial impressions are necessarily correct. My initial comment was triggered by the fact that it seems to me that most people commenting the decision seem to care only about Article 84 (which admittedly was the main focus in the proceedings leading up to the decision). However, r. 36-39 must mean something. In r. 41 the EBA states that it expects inconsistencies with certain EPC requirements rarely, if ever, to arise; the patentability requirements do not figure among those provisions mentioned "for the sake of completeness".
That is all I wanted to say. I am still trying to sort out my thoughts, so I am grateful to you and anybody else that might want to provide their views.
Proof of the pudding-09.09.2026 at 09.50.00 GMT+1
DeleteI you do not see why a disclosure relating to a different embodiment, i.e. lacking a limiting feature found in the claims, would lead to non-compliance of the claims with Art 56, then we have a problem.
Do you really think that if a claim is limited, a broader claim is still encompassed by the limited claim? If the original claim A+B has been limited to A+B+C due to relevant prior art, do you really think that the subject-matter A+B corresponds to the definition of the matter for which protection is sought?
I have always been taught that if a more narrower claim is lacking inventive step, any broader claim is even more lacking inventive step. The same actually applies to a lack of novelty.
In Reasons 5.6 of T 1391/24, the board held that claim 1 of AR3 as maintained by the OD was lacking IS. In Reasons 6 of T 1391/24, the board held that claim 1 as granted as well of AR1+2 being broader in scope, the lack of IS in claim 1 according to AR3 applies, mutatis mutandis, to claim 1 as granted and to claim 1 according to AR1+2. The same reasoning can be found in Reasons 25 of T 1018/23.
With a mind willing to understand, G 1/25 leaves, in my opinion, no real doubts as how it is to be applied. I am looking forward to the first board’s decisions applying G 1/25.
Like most practitioners you would have liked the EBA to abolish once for all the necessity of adapting the description after amendment of the claims. The EBA was not prepared to follow this route. Whether we like it or not is irrelevant. We will all have to live with it.
The more I mull it over, the more I think that it was the pemetrexed case that drove the Board to its conclusions here. In the USA the DoE is being crimped back and over here many think that the courts in pemetrexed should not have found infringement by equivalent. Never again, thought the EBA, shall the holder of a European patent be able to get to such an advantageous outcome
DeletePassing By, the way I see it, the EBA is saying in r. 38 and r. 39 that there is an Article 56 problem if the "real doubt" caused by the inconsistency leads to a claim interpretation which encompasses non-inventive subject matter.
DeleteReally, the main problem here is the nonsense definition of "inconsistency" in r. 18. As far as I can tell, the EBA is essentially trying to say that there is an "inconsistency" if, when interpreting a claim in accordance with G 1/24, the skilled person is left in "real doubt" as to the meaning of the claim. Forgive my bluntness, but that is a clarity test in all but name. It makes no difference whether the lack of clarity stems from the words of the claims or their interpretation in the light of the description.
A lack of clarity obviously has implications for compliance with provisions of the EPC other than Article 84. There is a whole body of case law which addresses this scenario - ie case law relating to oppositions in which Article 84 was not available as an objection against an unclear claim (see, for example, the GL at F-III, 12). What matters in that scenario is the subject matter that the (unclear) claim is interpreted to encompass and whether that subject matter complies with the provisions of the EPC. I think that it is no coincidence that the EBA's reasoning a r. 38 and r. 39 points to an identical approach.
@ Anonymous-10.09.2026 at 08.45.00 GMT+1
DeleteYour guess seem to be the right one. It is indeed, and I have said this before, the tightening of the screws in matter of adaptation of the description started with the famous pemetrexed case. In this case, the then applicant, insisted on keeping in the description a generalising statement, although the only salt ever disclosed was pemetrexed disodium. With G 1/25 this would not have happened.
What the EBA is after in G 1/25, is exactly what you conclude: the holder of a European patent should not be allowed to have a generalising statement in the description without being supported by any evidence. The generalising statement was actually infringing Art 123(2) and should have been removed.
Proof of the pudding-10.09.2026 at 12.40.00 GMT+1
DeleteLet’s start with the assumptions that the claims are clear as such. Only if a claim is clear as such can it be decided whether its subject-matter is novel and/or inventive. If the claims are not clear as such, then an objection of lack of novelty or inventive step can in general be raised on the basis of the broadest technical interpretation of the claim. Rather than arguing at length about the clarity of the claim, raising an objection of lack on novelty or inventive step is, for an examiner, the best way to tackle the claim and force the applicant to move.
Reasons 38 and Reasons 39 deal with inventive step. As far as inventive step is concerned, the doubt always benefits to the applicant. When it comes to inventive step some embodiments are inventive and some are not. I have never seen a decision from a board in which an embodiment is considered half inventive. There cannot be any real doubts whether the subject-matter of claim is inventive or not.
If some non-inventive embodiments are kept as filed in the description, then, there is a real doubt about which embodiments are inventive and supported by the description and which are non-inventive and hence not supported by the description. The hiatus or inconsistency between claims and description has to be removed.
This is what G 1/25 requires and to me, it is abundantly clear what constitutes an inconsistency: any statement in the description which allow to dilute the definition of the matter for which protection is sought. In such a situation, there are real doubts about what actually falls under the claim and what does not. What is at stake is not the clarity of the claim as such, but the clarity of the support of the claims by the description.
This does not hinder a jurisdiction deciding in post grant or post maintenance about the fate of patent whether there is infringement, be it directly, indirectly or by equivalence.
Just one question: why would a statement in the description which does not give rise to any real doubt about the meaning of the claims have any impact upon the assessment of inventive step?
DeleteI was under the impression that it was the claims which need to meet the requirements of Article 56 EPC, and not the description (and any drawings).
@ Proof of the pudding-11.09.2026 at 10.26.00 GMT+1
DeleteYou must have misread or misunderstood what I have said.
Indeed, it is the subject-matter of the which needs to meet the requirements of Article 56, and not the description (and any drawings), but you have conveniently forgot that according to G 1/15, it is by now abundantly clear that the claims have to be supported by the description and vice-versa.
I also notice that you have conveniently avoided replying to my comment citing T 1391/24 and T 1018/23. I wonder why?
The subject of the thread upon which you decided to comment has been r. 38 and r. 39, which relate to "inconsistencies" in the context of Articles 52 to 57 - and inventive step in particular.
DeleteNeither Article 84 EPC nor the support requirement is mentioned in either r. 38 or r. 39. So why bring that into the discussion now and claim to have always been referring to that provision?
Of course, it does not take a genius to work out why you decided to move the goalposts when confronted with facts that you cannot dispute. It also does not take a genius to work out why, when interpreting the EBA’s ruling, it is pointless to consider the reasoning of a decision (such as T 1391/24) that is neither cited nor discussed in G 1/25. It is therefore pretty obvious why I am not willing to waste time on such a futile endeavour.
The issue of support can be left for another day. The EBA’s ruling is nowhere near clear enough to determine the precise meaning of their rather cryptic comments in r. 35. I will say this, however: it would be rather surprising if, having gone to the trouble of defining the meaning of “inconsistency” in r. 18, the EBA decided to use a different (and unexplained) meaning of “inconsistency” in r. 35. Especially as r. 35 starts with “However, applying the above-defined concept of inconsistency”.
To my mind, this points to the requirement for an “inconsistency” to give rise to a lack of clarity of a claim (ie real doubts as to the meaning of the claim), even when the resulting objection is to a lack of support under Article 84 EPC. I do not expect you to agree. Indeed, I myself think that the EBA’s definition of “inconsistency” is nonsensical. But the above conclusion is nevertheless the only logical deduction that it is possible to make.
@ Proof of the pudding-11.09.2026 at 17.8.00 GMT+1
DeleteI never claimed that Art 84 is mentioned in Reasons 38 and 39. This is what you actually did in your comment of 10.09.2026 at 12.40.00 GMT+1: “that is a clarity test in all but name. It makes no difference whether the lack of clarity stems from the words of the claims or their interpretation in the light of the description.” You have changed the goal posts, not me. The same goes for T 1391/24.
I agree that T 1391/24 has not been cited in G 1/25 and has a priori no reason to be associated with the latter. I have cited this decision to show that your assertion that a claim “lacking a limiting feature found in the claims, would [not] lead to non-compliance of the claims with Art 56”. You might not be willing to waste time on such a “futile” endeavour. The endeavour is by no means futile as it shows that your postulate is manifestly wrong. You apparently cannot acknowledge that you have been caught on the wrong foot.
When reading the order of G 1/25, Art 84 is cited among plenty of other Articles of the EPC.
I agree with you that “the requirement for an “inconsistency” to give rise to a lack of clarity of a claim (ie real doubts as to the meaning of the claim), even when the resulting objection is to a lack of support under Article 84 EPC”.
If the description allows to consider embodiments not covered by the claims to still represent the matter for which protection is sought, it boils down in the end to a lack of support of the claims by the description, be it, inter alia, under Art 56, 83 or 84.
I have tried to use reason and logic to establish the legal foundations of your interpretation of G 1/25. I give up. Your reasoning makes no sense to me and it has become abundantly clear that your interpretation of G 1/25 was always going to be the same no matter how the EBA decided that case.
DeleteTo illustrate: can you think of any way in which G 1/25 has forced you to change your views on when (and how) a description should be adapted to amended claims?
Comparing the EBA’s rulings in G 1/24 and G 1/25 to the arguments in your amicus briefs, it is hard to spot much, if anything, in the way of substantive alignment with respect to either how the EPC should be interpreted (G 1/24) or the reasons for and consequences of the EBA’s interpretation (G 1/25). And yet, despite your views having previously relied upon arguments that have now proven to be almost entirely unfounded, it seems that you are claiming complete vindication. You will forgive me if I struggle to understand how you manage to reach that conclusion.
As for your comments on Reasons 5.6 of T 1391/24, those are based upon a fundamental misunderstanding. My comment concerned a disclosure relating to a different (broader) embodiment than that defined in the claims. On the other hand, Reasons 5.6 of T 1391/24 relates to different claims - of a broader claim request.
There is an important difference between the subject matter of a claim and the subject matter of a disclosure of the description. Whilst you often appear to conflate the two, Reasons 10 (1st sentence) of G 1/25 explains why they are very much not the same.
@ Proof of the pudding-14.09.2026 at 10.04.00 GMT+1
DeleteI am not bothered whether my reasoning makes sense to you or not. This is not my problem. You might appear under a pseudo, but it happens that to me, it is perfectly clear who you are and which was your amicus curiae for G 1/25. No adaptation of the description whatsoever. Reading G 1/25, and especially Reasons 21, 27-35, 36-39 and 40-41, I have “real doubts” that you can cling on Reasons10, 20, first sentence and Reasons 21, in spite of all what you have said in your amicus curiae.
You are manifestly thinking that G 1/25 has opened a new era in which the description does not need to be amended. I fear that you are on the wrong track.
I fail to see how my arguments have been proven to be almost entirely unfounded. I do not claim complete vindication, as the EBA has said that complete alignment between claims and description is not a matter of form but of substance. Reasons 21, 27-35, 36-39 and 40-41 are all about the substance and not the form. And those reasons have been summarised in the order of G 1/25.
As far as T 1391/24 is concerned, I also fear that you are fundamentally misunderstanding of what this decisions says. Your are playing with words in trying to make a distinction between the “disclosure” and the claims. A “broader disclosure” which bears only some relationship with the subject-matter for which protection is sought cannot be held to support the claims as required by Art 84. This is what can be read in G 1/25.
If you insist on the “broader” disclosure, then let us apply G 1/24. If the broader disclosure does have to have any meaning, then it has not to be read virtually into the claim, but has to replace in the claims the previous definition of the matter for which protection is sought, and you might end up, like in T 439/22-2 with a lack of novelty.
The original description can be compared to the foundations of a house. When the claims are narrowed during proceedings, the foundations must be adjusted to be in line with the definition of the subject-matter for which protection is sought. The claims can be compared to the part of the house above ground. Leaving the foundations unchanged if the part above ground has been reduced during proceedings would be tantamount to building a rather rickety house.
Art 69 and the Protocol, allow a court, when ruling after grant or maintenance in amended form, to consider that other windows, doors or the shape of the roof might also be belonging to the above-ground part of the house. This does not fall within the remit of the EPO. However, the foundations of the house must correspond to the part of the house above ground, i.e. to the claims.
When you come up with a “broader disclosure”, you seem to consider that the description can also be about a large shed which has only some relation with the house. How can the foundations of this large shed be in accordance with the foundations of the limited claim? I certainly do not conflate both, but I simply say with such a “broad” disclosure, the limited claims are not supported by the broad foundation of the shed. Conclusion: the house is anything but stable. and will collapse under the slightest wind gust=relevant prior art.
In the present blog, I found this revealing statement: “Patentees will always bring bad claims for as long as we have the protocol to article 69. Whether the description is conformed is actually irrelevant.” I see in this statement a clear malicious intent. Come up with a bad claim, stuff the description with lots of other things, and in case of litigation, play around with Art 69+Protocol. This allows to interpret the claim as you think fit. In other words, let’s have a go at making an magnificent Angora cat.
The duty of the EPO is to deliver solid patents, not patents for which the proprietor can decide on the interpretation as he thinks fit. This is what has to be taken from G 1/25.
Can someone explain to me: if DXT's understanding of G 1/25 is right, why did the Enlarged Board need to include so many caveats to define a legally relevant "inconsistency"? Under DXT's view, an inconsistency seems to arise almost automatically whenever the allowable claim scope is narrower than the originally filed application, regardless of whether there is any genuine possibility of misunderstanding the claim language with the unamended specification.
DeleteMost commentaries take a far more nuanced view of G 1/25 than the rigid approach DXT has been promoting for years. His approach seems heavily driven by a concern over how national courts or the UPC might interpret claim scope post-grant. However, forcing strict description deletions at the EPO will not restrict national courts in exercising Article 69 EPC and its Protocol to determine the extent of protection and assess equivalents.
In Pemetrexed, which often seems to be the elephant in the room, amendments were made to the claims during prosecution, yet none of the national courts evaluating equivalence were hampered by description differences. Across Europe, established criteria for the doctrine of equivalents focus on the claimed invention, the variant, and whether the variant performs substantially the same function in substantially the same way, not on whether the EPO forced a deletion of non-claimed subject matter from the specification. Furthermore, file wrapper estoppel is applied very sparingly by national courts, typically only when an explicit waiver was made to overcome prior art.
If the underlying concern is that a national court might grant protection based on equivalence to subject matter that was known or non-inventive at the priority date, the well-established Formstein's (or Gillette) defense already prevents that exact outcome across national jurisdictions and the UPC.
For these reasons, I don't see how the EPO's former strict approach to description adaptation would actually lead to a different post-grant outcome in court. If anyone can point to a case where it did, please let me know.
@ Proof of the pudding-14.09.2026 at 10.04.00 GMT+1
Delete@ Proof of the pudding-14.09.2026 at 10.04.00 GMT+1
Part 1
I am not bothered whether my reasoning makes sense to you or not. This is not my problem. Furthermore, I did not fall for any fundamental misunderstanding.
You might appear under a pseudo, but I have the privilege to appear under my name, so it was easy for you to see my amicus curiae brief. So what?
Reading G 1/25, and especially Reasons 21, 27-35, 36-39 and 40-41, I have “real doubts” that you can cling on Reasons10, 20, first sentence and Reasons 21, in spite of all what you might have said in this and in the parallel thread on G 1/25.
You are manifestly thinking that G 1/25 has opened a new era in which the description does not need to be amended. I fear that you are on the wrong track.
I fail to see how my arguments have been proven to be almost entirely unfounded. I do not claim complete vindication, as the EBA has said that complete alignment between claims and description is not a matter of form but of substance. Reasons 21, 27-35, 36-39 and 40-41 are all about the substance and not about the form. And those reasons have been summarised in the order of G 1/25.
As far as T 1391/24 is concerned, I also fear that you are fundamentally misunderstanding of what this decisions says. You are playing with words in trying to make a distinction between the “disclosure” and the claims. A “broader disclosure” which bears only some relationship with the subject-matter for which protection is sought cannot be held to support the claims as required by Art 84. This is what can be read in G 1/25.
If you insist on the “broader” disclosure, then let us apply G 1/24. In as far as the broader “disclosure” bears a relationship with the subject-matter for which protection is sought, it cannot be ignored. In other words, if the broader disclosure does have to have any meaning, then it has to be read into the claim, i.e. it has to replace in the claims the previous definition of the matter for which protection is sought.
The broader disclosure gives thus necessarily raise to real doubts about said definition, and we have an inconsistency in the meaning of the order in G 1/25. The broader disclosure can then be faced by broader prior art, whilst the limited claim cannot be attacked. You might then end up, like in T 439/22-2, with a lack of novelty.
I would not delete the broader disclosure, see T 439/22-2, but said broader disclosure cannot remain in the description as representing the definition of the matter for which protection is sought, i.e. the limited claim.
The original description can be compared to the foundations of a house. When the claims are narrowed during proceedings, the foundations must be adjusted to be in line with the definition of the subject-matter for which protection is sought. The claims can be compared to the part of the house above ground. Leaving the foundations unchanged if the part above ground has been reduced during proceedings would be tantamount to building a rather rickety house.
@ Proof of the pudding-14.09.2026 at 10.04.00 GMT+1
DeletePart 2
Art 69 and the Protocol, allow a court, when ruling after grant or maintenance in amended form, to consider that other windows, doors or the shape of the roof might also be belonging to the above-ground part of the house. This does not fall within the remit of the EPO. However, the foundations of the house must correspond to the part of the house above ground, i.e. to the claims.
If the claim is limited and the description refers to a broader disclosure, then there is an inconsistency. All your argumentation on T 1391/24 fails thus to convince.
Jurisdictions acting in post grant or post maintenance, are not bound by decisions of the EBA. They can made the interpretation they wish, but the EPO and the boards are. They will not ignore G 1/24 and G 1/25.
When you come up with a “broader disclosure”, you seem to consider that the description can also be about a large shed which has some relation with the house as limited by the claim. How can the foundations of this large shed be in accordance with the foundations of the limited claim?
I certainly do not conflate claims and disclosure, I draw logical conclusions based on G 1/24 and G 1/25. I simply say that with such a “broad” disclosure, the limited claims are not supported by the broad foundation of the shed. Conclusion: the house is anything but stable, and will collapse under the slightest wind gust=relevant prior art.
In the present thread, I found this revealing statement: “Patentees will always bring bad claims for as long as we have the protocol to article 69. Whether the description is conformed is actually irrelevant.” I see in this statement a clear malicious intent. Come up with a bad claim, stuff the description with lots of other things, and in case of litigation, play around with Art 69+Protocol. This allows to interpret the claim as you think fit. In other words, let’s have a go at making an magnificent Angora cat.
The duty of the EPO is to deliver solid patents, not patents for which the proprietor can decide on the interpretation of the claims as he thinks fit, under the pretext that the “disclosure” is broader. This is what has to be taken from G 1/25 and G 1/24. I maintain thus that G 1/25 is the continuation of G 1/24.
Mr Thomas, thank you for clarifying that you think that I am “playing with words in trying to make a distinction between the “disclosure” and the claims”. This helps me to understand why you are quite so insistent upon adaptations of a description which does not precisely align with the claims – and also why you have such a problem with “claim-like” clauses.
DeleteIt is clear to me that we have very different understandings of what the EBA meant in G 1/24 when it stated that “The claims are the starting point and the basis for assessing the patentability of an invention”.
For me, that statement in G 1/24 provides a complete explanation for the EBA’s later clarification (in Reasons 10 of G 1/25) that “The description and drawings may affect the meaning which the skilled person attributes to the claim wording, but they cannot be used to impose on the claim a limitation or expansion for which the claim wording provides no basis”.
I am not sure what that statement in G 1/24 means to you, or how you square it with your view that a broader disclosure in the description “has to replace in the claims the previous definition of the matter for which protection is sought”. Perhaps you need to stop thinking about this in terms of foundations for a building. It might be more apt to think about the claims as defining a structure to be built and the description as comprising a library of detailed building instructions. If the claims say “shed”, then the skilled person knows not to build a brick house – and will not have any “real doubt” about that fact even if the description (also) provides detailed instructions for building precisely such a house.
Just a thought.
A few more thoughts.
DeleteThe back-and-forth between us started with you challenging my comment (to Passing By), which was:
“Frankly, I do not see why a disclosure relating to a different embodiment (ie lacking a limiting feature found in the claims) would lead to non-compliance of the claims with Article 56 EPC. Can you?”
On 9 Sep, you pointed to Reasons 5.6 of T 1391/24 as supporting your view that a broader disclosure in the description (vs. that in the claim) would lead to non-compliance of the claim with Article 56 EPC.
So, to be absolutely clear, we were debating the standard for compliance with Article 56 EPC.
I therefore cannot help but draw your attention to yet another instance where you have moved the goalposts. Rather than addressing the point about Reasons 10 of G 1/25, you attempt to allege that I have fundamentally misunderstood what T 1391/24 says ... on the grounds that “A “broader disclosure” which bears only some relationship with the subject-matter for which protection is sought cannot be held to support the claims as required by Art 84. This is what can be read in G 1/25”.
Article 84 EPC and Article 56 EPC are two very distinct requirements. Compliance with each of those requirements must be assessed separately. As I said before, the issue of support can be left for another day. I am interested solely in what the EBA had to say about the issue of compliance with Article 56 EPC.
In this respect, if you cannot justify your position on Article 56 EPC without somehow relying upon the support requirement of Article 84 EPC, can I take it that you do not actually have a serious answer to the question that I originally posed to Passing By?
@ Anononymous-15.09.2026 at 10.58.00 GMT+1,
DeleteI first want to observe that the EBA has not included “so many caveats to define a legally relevant "inconsistency"”. The bulk of the decision and the order are abundantly clear. The scope of the claim is defined under Art 69 and the Protocol. G 1/24 has made clear that those requirements are not applicable in procedures before the EPO.
An inconsistency arises whenever the definition of the matter for which protection is sought is narrower than the originally filed application. If you apply G 1/24, if the claim might be narrower than the disclosure in the original description, then according to G 1/25 and T 439/22-2, the broader interpretation in the description replaces the one in the claim, and a novelty or inventive step objection can be raised. In T 439/22-2, the broader definition in the description was novelty hit by a broad piece of prior art in spite of the limited claim..
If you bank on a genuine possibility of misunderstanding the claim language with the unamended specification, then the least to be said is that either the description or the claim is shoddy. A claim ought to be clear and supported by the description. In G 1/25, the support is not formal but in substance. There is no caveat about this.
I am not, and I have never been in favour of deletions of parts of the description. They can be kept, but the situation has to be clairified.
The aim of the EPO is not to restrict national courts in exercising Art 69 and its Protocol to determine the extent of protection and assess equivalents, but to avoid the proprietor to change its interpretation as he thinks fit.
Whether in the pemetrexed case national jurisdictions have decided to infringement by equivalence is irrelevant as far as the procedure before the EPO is concerned. However, it happens that in this case, the original disclosure was limited to a specific salt, disodium. There was no way to bring in equivalents.
The then applicant wanted to claim pemetrexed in general, which was greeted by an objection under Art 123(2). The then applicant nevertheless maintained in the description a generalising statement. Under G 1/25, this will not be possible any longer, but it will not hinder national jurisdictions to decide infringement by equivalents.
It is pretty manifest that most commentators would have liked the necessity to adapt the description to be set aside. The EBA has not decided this way. I do not see a priori that the practice of the EPO and the boards will substantially change. The only change is that adaptation will have to be better argued from the side of the EPO and of the boards before they remit.
It is symptomatic that all commentors wanting a relaxed view of the EPO when it comes to adaptation of the description remain anonymous or write under a pseudo. I wonder why.
A B1 publication contains no abstract.
ReplyDelete@ Passing by-09.09.2026 at 19.06.00 GMT+1
ReplyDeleteWhen in G 1/25 the EBA says that, according to G 1/24, a broadening or restrictive feature has to be read in the claim, it does, in my opinion, not mean the definition found in the description is merely “virtually” or “intellectually” read in the claim and nothing changes.
What the EBA said, may be not in so many words, is that the definition found in the description is actually replacing in the claim the original wording. In doing this exercise, the inconsistency stemming form the different definitions in the claim and in the description is resolved. Then, the definition of the matter for which protection is sought is aligned in the description and in the claim.
In T 439/22 the board “wrote” the broadening definition in the description into the claim, with the consequence that claim 1 as granted lacked novelty.
When it comes to limiting statements in the description, only a general limiting statement in the description has to be “read”=”written” in the claim according to G 1/24. Not a limitation only present in one or more embodiments.
If the limiting feature is only found in one or more specific embodiments, G 1/24 does not apply and the limiting feature of a specific embodiment will be ignored, unless it is actually incorporated into the claim and replaces the broad definition in the latter. This is what most decisions applying G 1/24 have done. Those decisions are perfectly aligned with G 1/24 as it has been clarified in G 1/25. I refer here for example to T 2027/23.
The applicant/proprietor has the choice to align the claim with the description or the description with the claim, but the inconsistency has to be resolved. It cannot remain.
The applicant/proprietor cannot avoid, by drafting a limiting claim to be confronted with broader prior art, if the definition of the matter for which protection is sought is broader in the description. This is what happened in T 439/22-2.
The aim of G 1/24 and G 1/25 is to actually pin down the applicant/proprietor to one and the same definition of the matter for which protection is sought in the description and in the claim. This avoids the proprietor to change its mind with respect of interpretation of the claim depending on the circumstances.
I don't think I agree with you here. A limiting definition might very well also affect one embodiment and should be taken into account when assessing patentability, just like it is taken into account when assessing the scope. Same as for a broadening definition. E.g. claim 1 is about an antibiotic for treating humans or animals. The definition in the description for humans includes all hominidae (thus broadening if you will as it also includes certain human-like apes) and the definition for animals excludes insects and molluscs (thus limiting). I would not say that only the broadening should be taken into account for such a claim (i.e. disclosure of using the antibiotic for treating insects or apes would be novelty destroying, the latter also when the claim is limited to humans), but rather that also the limitation of only one embodiment should be taken into account if it is decisive (i.e. only prior art of use in apes is novelty destroying). Otherwise it does not make sense. The limitation in the description does affect the overall scope of protection of the claim, why should it thus not be taken into account for patentability? (this is what G1/24 is all about) Only because it affects only one embodiment that is encompassed by the claim?
DeleteOtherwise, if the applicant takes the definition into the claim, this might very well be contrary to Art. 84 EPC as well and a requirement that I think has been overlooked so far: claims have to be concise. And this for a good reason. So next time an examiner asks to have a definition in a claim I will just put a reference to the description in the claim to also fulfill this requirement. Otherwise we have a perfect Catch-22 situation: not being allowed to include a definition into a claim, but the claim is not patentable without it. Luckily R. 43 (6) EPC allows this.
@ BF-11.09.2026 at 09.13.00 GMT+1
DeleteI have said it before, and Max Drei agrees with me, that hypotheticals can lead to strange conclusions, as in the vast majority in the cases they relate to very specific situations,
What is important here is to discuss matters in a form as general as possible or applicable to the vast majority of the cases. Exceptions confirm the rule is a well known proverb.
It can thus well occur that some limitation or broadening linked to a very specific embodiment can lead to an objection or has to be taken into account. However, this will be more an exception than the rule.
What the bulk of the boards have said when applying G 1/24 is that a broad claim cannot escape a broad piece of prior art just because there is a limiting feature in one ore more specific embodiments disclosed in the description. As a general rule this is perfectly logical and should not lead to lengthy discussions.
You rightly mentioned R 43(6), but there again, from the wording of the rule alone, this situation will rather be an exception than the rule.
Regarding the abstract: the reason is simple why it is not taken into account: It is not necessarily written by the applicant/author. The patent office can change it at a whim and when filing an abstract it is merely a suggestion by the applicant that the patent office has to agree on and can even dispose with and write their own (see e.g. PCT R. 38.2).
ReplyDeleteAll the other issues/risks Rose mentions I might be wrong, but they where there before, so in my opinion nothing new under the horizon and no new risks (maybe only that now the examiners might really look into the description instead of this boiler plate language. So that Agfa/Gucci does not happen). While I was a candidate I was part of an opposition, where the applicant was deleting a definition and this was refused due to A. 123(3). We also once made an attack on a description amendment in the grounds for opposition (well, only to have the ground in the proceedings, but nonetheless we made it, which was not successful in the end). And this was all years before G1/24 and G1/25. But maybe this was back then a difference between biotech and mechanics?
At least also from a German background the approach of the UPC regarding the limited taking into account of the prosecution history does not seem so new. Only that it is now not restricted to Germany.
Also, wasn't all this not already argued in several amicus briefs that these are the risks of the current/past practice?
Where I agree with is that it might happen now that more arguments might be exchanged between applicant and examining division. And of course this might influence (as before G1/25) litigation. So in my opinion the quality of risks did not change. Maybe the quantity, but this we have to wait and see.
Regarding the consideration of a file prosecution estoppel by the UPC, I do not agree with Rose.
ReplyDeleteIn LD Milan Agathon v Intercom of 5 May 2026, the court noted that the claimant gave a different reading to a claim element than during prosecution, stating : “In view of the legal certainty, the complete turnaround in the statements made by the applicant/proprietor must be taken into account in the assessment”. The court referred to the case law of the CoA (20 December 2024, Alexion v. Amgen and Alexion v. Samsung Bioepis ; 3 October 2025, Beylkin v Philips), which explained that a statement during grant proceedings with respect to the meaning of a claim element can be seen as the meaning of such claim element for the skilled person at the time of the patent application. In these CoA decisions, the acknowledgment of the estoppel was implicit, it is now very explicit in the Agathon decision.
Regarding G 1/25, I find it positive. It seems it strengthens the burden of proof on the ED/OD to justify a request for description amendment, by referring to a « real doubt ». It is also positive that the « inconsistency » if any has to be assessed as part of the assessment of the EPC conditions, no longer at the 17(3) stage. In adition, if there is an inconsistency which makes it unclear whether an disclosed embodiment falls within the scope of the amended claim or not, applying the broadest sensible interpretation principle of the case law will be effective to urge the applicant to cure the inconsistency, and the logical way for that is to amend the claim so that it does not cover the embodiment any more.
I fully agree with Rose as to the concern about the risk of Art 123 issues resulting from description amendments. An ED should never suggest an amendment, be it in the claims or the description, without making sure that the amendment cannot create Art 123 issue. This is a duty, as pointed out in the SQAP section of the 2025 quality report. The SQAPs findings on granted patents include a consensus of the panel about the need to « check basis in original applications before proposing amendments ». Failure in this respect is therefore a serious quality issue. This duty adds to the burden of proof on the ED. Obviously, the EPO cannot allow itself to create legal uncertainty detrimental to the applicant’s detriment, especially when its policy objective is legal certainty.
The applicant is solely responsible for the content of a granted patent, the EPO has no liability, even in iatrogenic revocation cases such as the notorious T 1473/19. This gives applicants a strong argument for resisting a request for amending the description.
Mr Hagel-10.09.2026 at 11.49.00 GMT+1
DeleteI don’t think that in the future there will be more problems with Art 123 as we have had up to now.
It is difficult to follow that deleting one of two alternatives being not any longer covered by the definition of the matter for which protection is sought, when the claim has been limited, would imply an infringement under Art 123.
It is difficult to follow that the acknowledgement that some originally disclosed embodiments are not any longer covered by the definition of the matter for which protection is sought, when the claim has been limited, would imply any infringement under Art 123. In this case there is no need to delete the non-patentable embodiments, but to simply acknowledge that they do not fall under the limited claim. I have never been for a mere deletion of embodiments in this situation.
I agree that G 1/25 strengthens the burden of proof on the ED/OD to justify a request for description amendment. When, after limitation of the claim, the description still contains statements giving to the matter for which protection is sought a different interpretation than that derivable by the skilled person from the plain claim wording, there is a real doubt in the meaning of G 1/25.
I have always been against amendments to the specificationat the R 71(3) stage, not discussed beforehand with the applicant. G 1/25 contributes effectively in putting to an end this bad pratice of some EDs.
I agree that the EPO has no liability and it can only decide on documents the applicant/proprietor has submitted or agreed to. It is important to note that the “notorious T 1473/19”, has been issued well before the issuance of G 1/24. It is not to be followed as it explicitly refers to Art 69 when it comes to claim interpretation. I am however not convinced that under G 1/24 the outcome would have been much different.
In T 1473/19 the discussion might be have been about a “,”, but to qualify this revocation as iatrogenic goes a bit far. Iatrogenic is a term used in the medical field, and I have never seen it used in IP. It should not be forgotten that, in T 1473/19 the applicant approved the amendment introduced at the R 71(3) stage by the ED.
Interesting analysis. From a Turkish perspective this decision has direct relevance: Turkey's Industrial Property Code (SMK) Art. 89 requires claims to be construed with reference to the description and drawings, while Art. 92 requires that claims not exceed the scope defined in the description — an architecture quite close to Arts 69/84 EPC. As an EPC contracting state, Turkey is not formally bound by G 1/24 or G 1/25, but given the similarity of the underlying provisions, the decision will likely be treated by TÜRKPATENT and Turkish courts as persuasive comparative authority. Practitioners prosecuting the Turkish national phase should note that Turkish courts do actively rely on the description when construing claim scope in infringement and invalidity proceedings, so post-grant description consistency deserves the same care flagged here for EPO practice.
ReplyDelete