G 1/25, issued today, holds that the description must be adapted only where an “inconsistency” between the claims and description causes non-compliance with another requirement of the EPC. Legal basis for adaptation of the description, says the Enlarged Board of Appeal (EBA), can thus be found in the relevant EPC provision. The EBA defines our new word of the moment, "inconsistency", by reference to G 1/24.
The old divergence over legal basis of adaptation of the description is thus replaced by a new one over whether an incompatibility can be "readily" resolved or leaves the skilled person in "real doubt" (further choice phrases from the EBA). Critically, the requirement to amend the description must now be justified with reference to a relevant provision of the EPC such as novelty, inventive step, added matter etc. A patentee wishing to avoid make large-scale deletions of passages in the description may therefore be said to be better placed today than yesterday as amendment for amendment's sake is not considered justified. However, the price for this is that every deletion from the description now needs a reasoned justification on file. These reasons may bear on claim scope after grant. We may have to make fewer amendments, but these may be far more damaging post-grant.
In this Kat's view, the EBA has therefore changed the description requirement for the worse. The EBA has effectively changed it from something that could be dismissed as an annoying EPO formalities quirk to a substantive requirement for patentability that may consequently have far more impact on post-grant proceedings.
Legal background: A requirement in search of a legal basis
Article 84 EPC states
“The claims shall define the matter for which protection is sought. They shall be clear and concise and be supported by the description.”
Until now, the EPO has read into the second sentence of Art. 84 an obligation for the description to be brought into conformity with amended claims, either by deleting subject matter that is no longer claimed or by stating explicitly that it does not fall within the invention. However, the Boards of Appeal have been split as to the legal basis for this obligation. In one line of case law, the requirement of support was held to oblige the description to be consistent with the claims throughout, so that embodiments falling outside the claims must be deleted or clearly flagged. The opposing line of case law, running through T 1989/18, T 2194/19, T 1444/20 and, most fully, T 56/21, held that Article 84 EPC is a one-way street, i.e. it imposes requirements on the claims, not on the description, and that Rule 42(1)(c) EPC does not supply the missing basis.
T 56/21 came to the brink of a referral on this issue (IPKat) but then back-pedalled on the basis that the lack of legal basis was "unequivocal" and no referral was needed (IPKat). The referral eventually arrived anyway, from a case about hydroponics (IPKat), and the amicus briefs made it plain how seriously the profession takes this issue (IPKat).
 |
| Real doubt |
Case background
The referral came from T 697/22, an opposition appeal on EP 2124521, a patent for a hydroponics growing medium held by Knauf Insulation as Patentee and opposed by ROCKWOOL A/S as Opponent. The invention is a growing medium of mineral fibres held together by an organic binder. The Opposition Division had maintained the patent on the basis of the claims of auxiliary request 1E together with an amended description, both filed at the oral proceedings. Both parties appealed.
The Referring Board of Appeal found that the claims of the auxiliary request satisfied the EPC. Claim 1 of that request defined the binder with considerable specificity, requiring it to be based on reaction products obtained by curing an aqueous solution comprising citric acid, ammonia and dextrose, a definition imported from a dependent claim of the patent as granted and one that narrowed claim 1 substantially. At the oral proceedings, and for the first time, the Patentee filed a further amended description from which paragraphs [0013] and [0016], containing broader contradictory definitions of the binder, had been deleted. That description was not admitted, on the basis that it was late filed. The Referring Board of Appeal was thus left holding an allowable set of claims alongside a description that still contained two paragraphs defining the binder in terms that the amended claim no longer matched. The Board of Appeal identified an inconsistency between the amended claim and the offending paragraphs [0013] and [0016], and asked the EBA whether the EPC required that inconsistency to be removed, on what legal basis, and whether the answer differed in examination.
An inconsistency is whatever “consulting” the description according to G 1/24 cannot fix
The Enlarged Board of Appeal began by observing that neither the Referral Decision nor the EPC defines "inconsistency" (r. 17). It then supplied a definition:
“An inconsistency between the claims and the description, and any drawings, exists where one or more statements in the description, including the drawings, suggest an understanding of a claim which is incompatible with the apparent meaning of the claim, and that incompatibility cannot readily be resolved by applying the principles set out in G 1/24.” (r. 18, emphasis added)
The EBA then went on to explain that most apparent incompatibilities can be resolved without difficulty by applying the principles set out in G 1/24. What remains is an inconsistency:
“If the person skilled in the art reading the claim in the light of the description and any drawings would be left in real doubt as to the meaning of the claim, however, there is an inconsistency.” (r. 19, emphasis added)
So the test for whether a passage is inconsistent with a claim is now whether the incompatibility can be resolved by applying G 1/24, and, if it cannot, whether the skilled person is left in "real doubt" as to the meaning of the claim. Crucially, "readily" is not defined anywhere in the decision. Nor is "real doubt". As a Katfriend has noted, why “real doubt” and not “serious doubt”, a term already well used in the case law? For this Kat, the EBA has thus now built on the uncertainty of G 1/24 with yet more uncertainty. Boards of Appeal are now left with the question of how to determine whether inconsistencies are “readily resolved” and whether a skilled person would be in real, as opposed to presumably imaginary, doubt as to the meaning of the claim.
One point this raises, which has so far been ignored, is the question of the criteria by which real doubt is established for the skilled person, and particularly whether this takes account of what the skilled person may understand about the quirks of patent drafting. If a skilled person knows that patents include definitions that don’t read on to the claim, can they be said to be in real doubt with respect to the meaning of the claim in the face of such definitions? If the skilled person knows that the claims define the invention, can the description produce any real doubt at all?
Leaving this aside, the bar has certainly been raised for when description amendments are mandated. For anyone who has ever been asked to delete a perfectly good example:
“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, emphasis added)
Then, for applicants that wish to argue against description amendments per se, the EBA provides the helpful sentence:
“The EPC does not require an adaptation of the description, including any drawings, merely for the sake of formal concordance.” (r. 22)
Neither line of case law survives intact
The Enlarged Board of Appeal declined to adopt either camp of case law wholesale. The second line of case law, it held, rests on premises that "can no longer be maintained after G 1/24" (r. 28), because those decisions treat Article 84 EPC as assessable without reference to the description at all. On the one-way street argument, the EBA was clear:
“The Enlarged Board is of the view that a natural reading of Article 84 EPC, in particular the sentence, "They [the claims] shall be clear and concise and be supported by the description" cannot be said to impose a unilateral directionality that renders the description, including any drawings, irrelevant when inconsistencies compromise the understanding of the claims.” (r. 32)
The first line of case law fared no better. Article 84 EPC, the Enlarged Board of Appeal held, "does not require a purely formal concordance between the description, including any drawings, and the wording of the claims, nor does it impose a general obligation to remove from the description, including any drawings, all matter not reflected in the claims" (r. 34). What it does require, it stated, is that if it is unclear whether information, examples, subject matter or embodiments are or are not within the scope of the claim, the claim cannot be said to be supported by the description.
As such, the EBA concluded:
“It follows that the necessity to adapt the description or any drawings is not a consequence of the existence of an inconsistency as such, but arises only where, and to the extent that, the inconsistency has legal significance because it leads to non-compliance with a requirement of the EPC.” (r. 43)
Question 2 was accordingly answered to the effect that the legal basis for any necessary adaptation is the provision of the EPC with which compliance is lacking by reason of the inconsistency in question. Question 3 was answered "No": the interpretative role of the description does not depend on the procedural stage, so examination and opposition are treated alike (r. 13 to 15). The Order lists Articles 52 to 57, 76(1), 83, 84, 123(2) or 123(3) EPC as the provisions that might be engaged, whilst the reasons note that the possibility of Articles 83, 76(1) and 123 EPC being engaged is "more theoretical than practical" (r. 40) and that such issues will "rarely, if ever, arise in the future" (r. 41). It is interesting to see the inclusion of Article 123 EPC added matter as one of the provisions unlikely to be engaged.
What this means in practice
The question is thus no longer whether the EPC contains a general description adaptation requirement. The question is whether a particular passage in the description, left as it stands, causes non-compliance with a particular provision. An applicant resisting amendment can now ask the Examining or Opposition Division which provision is not complied with respect to the wording of the description, since r. 43 makes that provision the legal basis and there is no longer a free-standing tidiness requirement to fall back on. As such, it seems that generic requests to bring the description into conformity may now be pushed back on. It will also be possible to argue about the extent to which "real doubt" is introduced.
On the downside, however, all this may lead to more reasoned objections and arguments on file with respect to why passages of the description are deleted, potentially influencing claim interpretation in post-grant proceedings. There is also the concern that deleting passages from the description might be taken as an acceptance that the offending paragraphs altered claim scope for the purpose of assessment under novelty and inventive step.
Interestingly, the EBA gives a worked example (r. 38) in which a description passage expressing the technical teaching of the pre-amendment claim conflicts with the inventive step case built on the amended claim. Where a claim has been narrowed precisely in order to establish non-obviousness, the statements of advantage and technical effect in the description now need attention as a matter of Article 56 EPC, not merely as housekeeping. That is a more demanding review than the deletion exercise it replaces, and it cannot be delegated to a formalities check. The EBA has therefore made the work for applicants even harder, it seems, and the description adaptation requirement even more onerous.
Final thoughts
G 1/25 is built entirely on G 1/24. The definition of inconsistency in G 1/25 is expressed as a residue, being what is left after the principles of G 1/24 have been applied. On G 1/24, the EBA confirms the holistic approach to claim interpretation, endorsing T 439/22, and finding that the description and drawings may affect the meaning the skilled person attributes to the claim wording but cannot impose a limitation or expansion for which the claim wording per se provides no basis (r. 10), with relevance to the pending referral in G 1/26. Notably, however, the core issue in G 1/26, being the consequences for added matter of a lack of alignment between the description and the claims, is left unaddressed.
From G 1/25, therefore, the description adaptation requirement no longer floats free of the EPC and must now be pinned to an identified provision. An applicant or patentee resisting a request for wholesale deletion of passages now has EBA-level reasoning to argue against it. However, the consequence is that deletions made to the description may be given greater weight in post-grant interpretation of claim scope, in Europe and beyond. The number of amendments to the description may be fewer, but their potential adverse impact may be considerably more. Practically, this means more work for attorneys, greater costs for clients, and, in this Kat’s view, no substantive gain in legal certainty.
Further reading
- Board of Appeal poised on the brink of a referral on description amendments (T 56/21) (July 2023)
- Board of Appeal back-pedals on referral in view of "unequivocal" lack of legal basis for the description amendment requirement (T 56/21) (October 2024)
- Beware of boilerplate: Practical lessons for patent drafting from G1/24 (Claim interpretation) (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)
- G 1/25 (description amendments) amicus curiae: The battle lines are drawn (January 2026)
- 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)
Reviewed by Dr Rose Hughes
on
Thursday, September 03, 2026
Rating:
Follow the IPKat on LinkedIn 
![[GuestPost] What has long arms and lives in Switzerland? The cross-border injunction club gains a new member](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhbpnVQxI6tEyGQFH2BPgJnukkqAaRYeRNZvR9BdXHyH03ueEDFCCA9Tlmt1r_CtbVfkVfRGK37kXT4jyaWt5_qcBeRXnf2y2NT5yWc_l0fFsGG7AQpdFgTIGblfBeOMQaOJuyge1PR5cHWm7xxJbpqY7EVAYO1ouZacdexQ4mDx9dFhjnaVxp1/s72-c/Round%20Kat.jpg)
This decision may also open up new attack lines in opposition proceedings, as the opponent could argue that, e.g, in view of claim amendments, the (amended) description comprises an inconsistency. As late filed descrptions do not appear to be admissible, the patent proprietor may be forced to submit multiple auxiliary requests with the same claim set but different versions of amended descriptions to defend against different potential inconsistency objections. A real headache for the patent proprietor and potentially an easy way for the opponent to get a patent revoked based on a formality…
ReplyDelete@ Anonymous.03.09.2026 at 18.54.00 GMT+1
DeleteWhen starting an opposition, what has been granted cannot be attacked under clarity, cf. G 3/14.
G 1/15 does not open up new attack lines at the beginning of opposition proceedings. If the claims are amended, cf. G 3/14, an attack under Art 84 is possible in as far as there has been an amendment. This is however nothing new under the sun. I have seen this practiced in lots of oppositions. G 1/25 does thus not open new attack lines.
The question of an adapted description for each AR is however a real problem. It is only due to the fact that some boards are very academic, not to say pedantic when it comes to the application of the RPBA. This depends on the boards, but it is not a problem for opposition divisions.
A few initial, disconnected thoughts:
ReplyDeleteParagraphs 4-10 seem to very firmly put paid to the so-called "acontextual" approach to claim interpretation that has been clung onto by so many Boards and departments of first instance even after G 1/24. By explicitly endorsing the approach taken in T 439/22 this provides some welcome clarity to what "consulting the description" might mean. (It is notable that - as DXT has often been at pains to emphasise - the approach taken in the final decision in T 439/22 has up until now been an outlier, at odds with the majority of the other boards. Interesting to see that its approach now has the stamp of approval from the EBA.)
Paragraphs 18-22 seem to amount to a significant curtailing of the circumstances in which an "inconsistency" might necessitate amendment to the specification. I am not sure that I therefore share this Kat's pessimism - it seems squarely to place the onus on the ED/OD/BOA to justify why a statement is not only "inconsistent" but intolerably so in the sense that it cannot be resolved through, essentially, common-sense interpretation. Of course, the proof of the pudding will be in the eating - will the EPO try to find ways to finesse, or ignore, the EBA's ruling just as it did in G 1/24?
I don't see anything here that curtails the ability of the patentee to voluntarily delete or amend passages of the description, either, so again I'm not sure I share the Kat's pessimism about the downstream impact of amendments needing to be justified. Rather, my reading is that this decision significantly fetters the ability of the EPO to mandate amendments, while leaving the patentee substantially free to be the master of their own fate otherwise.
Initial thoughts, part 2:
ReplyDeleteThe reasoning on the bidirectional nature of Art 84 fails to convince (but, nevertheless, it has the force of law). I still fail to see a compelling reason as to why "support" is not given if only a part of the specification is consistent with the claim. Provided that at least some of the specification is consistent, what does extraneous information matter? Perhaps the reasoning here could have benefitted from being more detailed. Nevertheless, the EBA has spoken.
Can anyone tell me what paragraph 38 means? The Kat suggests that this relates to statements of advantage and technical effect. I am not so sure. If that's what the EBA meant, surely it would have said so? And isn't this problematic if the advantage or technical effect relied upon to establish an inventive step is one that is not expressly stated in the specification, but nevertheless permissible under the G 2/21 criteria? What if the claim has been narrowed and allowed without any express objections against, or arguments in favour of, inventive step being put on the record? Is there another interpretation whereby "technical teaching" in [38] is to be equated with "subject-matter" (e.g. in a sense analogous to the test under Article 123(2)) - but then in that case, is the implication that any subject-matter broader than that which has been found to be inventive, must be deleted? Is that not contrary to the reasoning in the preceding paragraphs?
Paragraphs 40-42 are highly disappointing. These seem to acknowledge the possibility that - as mentioned by several amici in their briefs - mandatory amendment could lead to inescapable traps under Article 83 and 123. And yet the Board merely handwaves these away as being unlikely to occur and mostly theoretical. This blithely disregards the fact that in the final decision in T 439/22 we have a concrete example of an inescapable Article 123 trap being created as a result of a mandatory description amendment based on (what the EBA now says) is the correct approach. The message from the EBA seems to be, "too bad - not our problem!"
For a decision which was taken by an EBA comprising Arnold, following proceedings where Arnold led the questioning, it is curiously brief and lacking in detailed reasoning. I wonder to what extent there might have been agreement behind the scenes on the overall conclusions but disagreements on how to get there and the implications, so that what we are left with is the bare minimum that the EBA members could agree on.
Even as someone who believes there is no basis for description amendments, and who finds r. 32 terribly conclusory (how can T 56/21 be dismissed with a single sentence of “reasoning”?), I am far more positive than you are on r. 38 of G 1/25.
ReplyDeleteThe way in which I read G 1/25 is that amendment can only be insisted upon where there is genuine rather than hypothetical ambiguity over whether a claim feature - interpreted in light of the description, of course - satisfies Article 54 or 56. This is to try to avoid the situation in T 439/22 where an unequivocal definition led to ambiguity on validity. This can only be the case where the examining or opposition division has an item of prior art and cannot decide whether it is within or without the claim scope.
Following the logic in G 1/25, the fact that feature X is not claimed is not a sufficient reason to delete it if it is obvious that X is not claimed. Presumably, then, the fact that the description says “Using X provides an efficient widget” is also not sufficient to have to delete it, as long as it is clear the claims do not provide the efficiency gain attributable to X because they clearly do not comprise X. While r. 38 is worded rather broadly, this result seems clear to me when r. 38 is read in conjunction with r. 37.
Hence, when the application contains consistory clauses and statements of advantage that are clearly unrelated to the claims, or where the application contains an obviously unclaimed example and data for that unclaimed example, there should not be any need to amend the description.
It is only when validity actually hinges on which way the claim should be interpreted in light of the description that amendment will be needed - and it may be possible to amend the claims to unambiguously exclude one interpretation to thereby fix the inconsistency without even touching the description. G 1/24 did tell us that the claims are the starting point and the basis for assessing patentability, after all.
Haven't yet read the Decision but wanted to get my comment in, early in the thread. Am I right, that the decision was written by David Rogers? If so, I am inclined to support Mouse's surmise, that the decision text reveals "the bare minimum that the EBA members could agree on".
ReplyDeleteBut is his necessarily something to be regretted. Is not this "bare minimum" something of high value"? Does it not set a welcome change of course and direction of travel?
To the contrary, Max, I think the "bare minimum" is to be regretted. That's how we arrive at impenetrable, inconsistent, incoherent and nonsensical ECJ rulings on SPCs, because - according to more than one insider I've spoken to - in the absence of a mechanism for dissenting views, what comes out at the end is a sanitised "minimum viable product" which leaves ample room for disagreement and interpretation among practitioners, patentees, and the wider public.
DeleteI think it would be far better if we accepted that there can be diverging views, even among the highest judges, and allowed them to express their reasoning in full so that the intellectual faultlines could be fully exposed and then explored in future cases.
I agree that this has the hand of Rogers all over it (not least because it seems mainly to be a decision which adds flesh to the bones of his preliminary opinion). I admit that, from personal experience, I have been far from impressed by any of his decisions, which - in my personal view - often smack of trying to get to a conclusion as quickly as possible, sidestepping any controversial issues, and to hell with rigour, legal justification, or consideration of wider implications.
I have been in more than one hearing where he, as the legal member and with little to do when the argument was mostly technical, appeared to be quite literally asleep. But perhaps others have different views. He also - again, in my personal view and experience - often seems to have little time for parties' practical circumstances, which sits very uncomfortably with his dismissal of Art.83 or 123 concerns as being mostly theoretical.
Yes. I read somewhere that it is the task of the highest appeal instance to bring clarity where there was a deficit of it (as opposed to the opposite). Have they doe that here. I doubt it.
DeleteI think that the combination of i) a holistic approach and ii) Art 84 cutting both ways and iii) conforming amendments needed also at the conclusion of post-grant proceedings, is going to make it more complicated, rather than less so, to apply for a patent at the EPO and carry it through to asserting it. But then, who else but corporate applicants cares? And is anybody going to shed a tear for their enforcement difficulties?
Well, the EPO didn't seem to want to expand on description amendments but they were pushed and they gave us this in return. So if you enjoy complaining about description amendments I'm sure this will fuel your engine.
ReplyDeleteI don't think this decision will change much in practice. Examiners might include some slightly updated boilerplate wording in OAs. I'm sure some will enjoy thinking up hypothetical ways in which it might significantly affect things but that never actually materialise.
Anon, I think that the chances are slim that the EPO will significantly change their description adaptation practice in the light of G 1/25. Frankly, the EBA's ruling is neither clear nor consistent enough to force the EPO to make the practice changes that are clearly necessary.
DeleteI have seen that an ardent proponent of description adaptation is already arguing that Reasons 35 of G 1/25 justifies current EPO practice. Of course, such arguments rely upon interpreting Reasons 35 entirely out of context, and ignoring explicit comments of the EBA which point to a very different conclusion. However, the EPO has form when it comes to (deliberately?) misinterpreting EBA rulings, with both G 1/24 and G 1/21 being cases in point. I therefore have no reason to expect that the EPO’s treatment of G 1/25 will be any different.
From the comments of a retired EPO Examiner on another blog, I gather that G1/25 might be used to ratchet up the severity of examination under Art 84 EPC even higher. The argument seems to be: The scope of the claim is less than 100% clear to me. I can't be sure with 100% certainty what acts it covers and which ones it does not.That is all the proof I need, that there exists an inconsistency between the description and the claim. This inconsistency is what is giving me "real doubt" as to what the claim covers. If you don't amend, to remove the inconsistency, I will refuse you.
DeleteG 1/25 beefs up GL F-IV.4.3 (2024 revision) which defines “inconsistency” as follows :
ReplyDelete“Any inconsistency between the description and the claims must be avoided if it casts doubt on the subject-matter for which protection is sought and therefore render the claim unclear or unsupported under Art. 84, second sentence, or, alternatively, render the claim objectionable under Art. 84, first sentence.”
The wording “if it casts doubt” of the 2024 revision replaced the wording “if it could throw doubt” of the previous version. It was indeed meaningful to replace a conditional by the present tense. This made it clear that as pointed out by Anon Y. Mouse, the onus is on the ED or OD to prove the need to remove an inconsistency. It was no longer possible to rely on hypotheticals as was possible under the “if it could..” language or on a mere reference to the Guideline devoid of substantiating reasoning.
The “real doubt” language of G 1/25 implies that the burden of proof will be stronger.
In addition, alongside the 2024 revision of the GL, the EPO’s practice of entering substantial amendments of the description at the 71(3) stage without prior consultation with the applicant was put to an end. G 1/25 pins another nail to the coffin by requiring that the issue of description amendment be handled as part of the assessment of the EPC conditions.
All in all, it seems to me the message is that substantial description amendments should not be considered except if there is a clear need based on substantiated reasoning. This would be wise for procedural efficiency and for avoiding 123(2) issues.
For me, this opens a Pandora's box. We are often forced to acknowledge the closest prior art in the description or the examiners even on their own introduce it.
ReplyDeleteG1/25 established no difference between the pre-grant and Post-gräbt proceedings. If Article 56 is the basis to acknowledge the closest prior art known at the pre-grant stage - would an unsecusfull opposition attack based on another closest prior art require the patent to be maintained with claims as granted but with an adapted description acknowledging this new prior art?
The EBA did not need to establish a difference between pre-grant and post-grant because the EPC already does. In opposition proceedings amendments are only necessary if they overcome a ground for opposition or if another amendment caused a noncompliance with the EPC.
DeleteThe basis for requiring the relevant prior art to be acknowledged is Rule 42(1)(b), which isn't listed as a ground for opposition, and I don't see how a regular amendment can cause noncompliance with R. 42(1)(b). I guess if the patent proprietor in opposition proceedings actively removes an acknowledgement from the granted patent, then that could be said to cause noncompliance with R.42(1)(b).
@ Jack Nicholson-07.09.2026 at 11.12.00 GMT+1
DeleteI cannot agree with your statement as you seem to have manifestly overlooked R 86.
According to R 86, “Part III of the Implementing Regulations shall apply mutatis mutandis to documents filed in opposition proceedings.”
Part III of the Implementing Regulations covers R 35 to 54, and hence also R 42(1,b).
Under R 86, R 50 relating to documents filed subsequently applies thus as well, for instance the claims have to be submitted on a separate sheet, e.g. for each auxiliary request.
The proprietor can remove an acknowledgement from the granted patent, only if this acknowledgement is replaced by an acknowledgement of the prior art having led to a limitation of the claim.
The proprietor cannot escape R 42(1,b).
@Anonymous
ReplyDeleteYou can never be forced to acknowledge the closest prior art in the description, if you mean by "acknowledge" a statement that the document is the closest prior art. You are even entitled to limit the amendment to a mere mention of the document, but without any comment as to the content of the document. A strong argument for refusing to comment on a prior art document is that any comment potentially violates Art 123(2) by altering the interpretation of the claims. Another strong argument is that Art 113(2) requires that any amendment be approved by the applicant.
Even mere mentioning of the document goes for me too far. What is the legal basis for this at the grant stage? What inconsistency is resolved by such an amendment?
DeleteAt grant stage the legal basis is R.42(1)(b).
DeleteOr rather: the inconsistency to be resolved is noncompliance with R.42(1)(b), and if you insist on not resolving it (the applicant always has the right to disagree with amendments made by the examiner), the application will be/should be refused under A.97(2).
The addition of a prior art statement was a topic discussed in detail in the IKPat post of 5/1/2022. But this was before the 2024 revision of the Guidelines.
DeleteUS practice is quite different but it is of note that US patent practitioners keep away from any comment in the IDS they are required to file, the IDS only provides a list of documents.
Someone needs to tell all the examiners I have then.
Delete@ francis hagel-06.09.2026-at 09.51.00 GMT+1
DeleteI refer to my reply to Jack Nicholson-07.09.2026 at 11.12.00 GMT+1.
You have at least to identify the document having led to the limitation of the claim. I refer here to the GL F-II, 4.3.
If your claim is in two part form, then a mere reference to the preamble of the claim is enough.
If your claim is not in two part form, then you have to enumerate all the features known from the document having led to the limitation of the claim.
Whatever solution is valid in the specific case, care should be taken to respect the considerations in the Guidelines H-V, 2 in general, and H-V, 2.4 in particular.
Art 113(2) is less strict s you make out: The EPO shall examine, and decide upon, the European patent application or the European patent only in the text submitted to it, or agreed, by the applicant or the proprietor of the patent. The proprietor can submit a text, but this does not mean that what is submitted or agreed complies with the requirements of the EPC.
I am a little confused by the EBA’s reasoning in connection with Article 84 EPC.
ReplyDeleteAt Reasons 33, the EBA makes it clear that the claim interpretation principles set out in G 1/24 are to be used not just for assessments under Articles 52 to 57 EPC but also for assessments under Article 84 EPC:
“The principles on claim interpretation set out in G 1/24 apply where the assessment under Article 84 EPC requires the meaning of the claims to be determined”.
However, this appears to directly contradict the EBA’s (admittedly obiter) reasoning in Reasons 6.2 of G 1/04:
“In the present context, it is further to be considered that Article 84 EPC requires that the claims define the subject-matter for which patent protection is sought, and that they must be clear. It signifies that an independent claim within the meaning of Rule 29 EPC should explicitly specify all of the essential features needed to define the invention, and that the meaning of these features should be clear for the person skilled in the art from the wording of the claim alone. The same should apply mutatis mutandis in respect of a claim relating to the subject-matter excluded from patent protection under Article 52(4) EPC. These requirements serve the overriding purpose of legal certainty”.
Does this mean that G 1/24 overrules this aspect of G 1/04, and that the clarity of a claim will now only be assessed once the subject matter of the claim has been determined by interpretation in the light of the description?
If so, this could represent a very significant change in practice. For example, it could make it impossible for the EPO to allege non-compliance with Article 84 EPC when an “unusual” definition of a term used in the claims is present only in the description.
At present, the GL at F-IV, 4.2 indicate that a claim must, as far as possible, be amended to include the “unusual” definition. The question that I have in the light of G 1/25 is: why?
I see that, over on another blog, DXT is furiously arguing that the EPO's previous strict practice has been fully vindicated and that anyone who disagrees is stupid and/or unfit to be a patent attorney (I wish I were exaggerating).
ReplyDeleteI wonder whether this gives us a foretaste of the way that this decision will be interpreted within the walls of the EPO? Somehow I think that the Enlarged Board has once again managed to find a "compromise" which merely stores up more arguments for the future.
@ Anonymous-08.09.2026 at 1402.00 GMT+1
DeleteIn my own blog, I have made clear that G 1/25 has not abolished the necessity of adapting the description. On the contrary it has enshrined it in examination and opposition.
In a reply to Anon-06.07.2026-2, I wrote “Embodiments which are not any longer encompassed by the amended claim in view of an objection of lack of N or IS, cannot be held to be supporting the amended claim as required under Art 84.
I have added “Is this so difficult to understand? If your answer is yes, you better look for a job outside the patent world.”
This is what G 1/25 has said whether you like it or not. If you do not like it, then there is a solution.
I take full responsibility for this words as they are the expression of my exasperation. You keep repeating the same arguments like a Tibetan praying mill. On top of it, you never even tried to discuss the examples I have given. The best example is the famous Agfa/Gucci case before the UPC CFI LD Hamburg. I did not hear a squeak coming from you about this example.
If in such a blatant case, the description does not have to be adapted, then the description will never have to be adapted.
As I have left the EPO 14 years ago, I express my personal views, and I would not be presumptuous enough to speak on behalf of the EPO.
Fantastic news.
ReplyDeleteHere we all were thinking that the livelihoods of patent attorneys were on the chopping block because AI was about to automate claim interpretation and analysis and patent prosecution more widely.
But then the Enlarged Board comes along with new law so far-reaching and poorly defined that you could not scarcely trust AI to apply it. And then the EPO (President, Guidelines) will get to reinterpret that Enlarged Board decision. They will presumably ignore any parts of it that do not support the status quo ante, and use selected quotes as confirmation of everything they have been saying all along. Cue lengthy back and forth between attorneys, worried about unnecessary amendments adding subject matter and affecting claim scope, and examiners, using G1/25 to justify sweeping amendments to the description.
Poor applicants...
Proposed solution: Article 84 EPC requires that the claims are supported (i.e. their scope justified) by the description (i.e. its technical teaching). If the claims are broader than justified by the description, Article 84 EPC may be invoked to require limitation. If the claims are narrower than justified by the description, the patent itself identifies unclaimed subject matter that works - great news for the public! And if there is some inconsistency between the claims and the description, decide what the claims mean! Sometimes, that will mean using a disclosure or a definition in the description to adjust the scope of the claim away from its strict literal meaning. And sometimes, it will mean ignoring that disclosure or definition and sticking with the strict literal meaning of the claims. That should be decided on a case-by-case basis based on technical considerations from the perspective of the skilled person.
Too controversial?
I am also confused about the EBA's definition of "inconsistency". According to Reasons 18 of G 1/25, an “inconsistency” exists where:
ReplyDelete(a) one or more statements in the description, including the drawings, suggest an understanding of a claim which is incompatible with the apparent meaning of the claim, and
(b) that incompatibility cannot readily be resolved by applying the principles set out in G 1/24.
But what is “the apparent meaning of the claim”?
Presumably the meaning of the claims (apparent or otherwise) is that determined by applying the principles of G 1/24. As described in Reasons 9, this involves “determining the meaning of the claim wording from the perspective of the skilled person based on the claims, the description and any drawings taken together”.
If this understanding is applied, an “inconsistency” exists where:
(A) one or more statements in the description, including the drawings, suggest an understanding of a claim which is incompatible with the meaning of the claim wording from the perspective of the skilled person based on the claims, the description and any drawings taken together, and
(B) that incompatibility cannot readily be resolved by determining the meaning of the claim wording from the perspective of the skilled person based on the claims, the description and any drawings taken together.
Is it me, or is (B) entirely superfluous here?
Also, in what universe does (A) make sense? If the meaning of the claim wording has already been determined by consulting the whole description, how could any part of that very same description suggest a different meaning?
And one more thing: what does the EBA mean by suggest an understanding of a claim? Does that require the incompatible understanding of a claim to merely be a possible interpretation based upon the description passage in question? Or does there instead need to be a fully justified interpretation that takes into account the full disclosure of the application as filed?
The former option seems unlikely, as it would require objections to be raised on an entirely hypothetical (ie inadequately reasoned) basis.
On the other hand, it is hard to see how an “inconsistency” could ever arise under the latter option – as that would involve comparing the understanding of the claims based upon the whole contents of the application as filed to ... the understanding of the claims based upon the whole contents of the application as filed!
I may be missing something, but it is hard to see how to make sense of Reasons 18 – unless all that it means is that amendment (of the claims or the description) is required in circumstances when an interpretation according to G 1/24 leads to the conclusion that the meaning of a claim is ambiguous.
Part 1
ReplyDeleteI have commented G 1/25 in my own blog and read with interest all the comments above. As I have even been mentioned twice, once directly and once indirectly, I have a further reason to comment.
G 1/25 contains a definition of an inconsistency: An inconsistency between the claims and the description, including any drawings, is relevant for the purposes of the EPC if, in the circumstances of the case, it leads to a non-compliance with one or more requirements of the EPC.
What I notice in this blog and in the comments, is that there seems to be a kind of fixation on the Reasons 18-20, and ignoring Reason 21 and more importantly Reasons 35 and 38.
The logical conclusion of Reason 35, is that the claim is not supported by the description, if the matter for which protection is sought has received a different definition, one in the claim and in one in the description.
What is at stake with G 1/25, is not the clarity of the claim as such, but the clarity of the claim in relation to the information found in the description. If the claim and the description tell a different story, the EBA has made clear that that the claim is not supported by the description and this inconsistency has to be resolved.
The logical conclusion of Reason 38, is that if the description contains, after any amendment of the claim, so as to render it allowable under Art 56, a statement, including any drawings, expressing a technical teaching reflecting the claim before this amendment, this statement conflicts with the fulfilment of the requirement of non-obviousness. There is thus an inconsistency in the description, including in the drawings. This inconsistency has to be resolved.
Reasons 40-42, cannot be held disappointing as they make clear that any amendment can bring about problems under Art 123. Just look at T 439/22, Reasons 5, where the deletion of § [0035] has been objected under Art 123(3). An objection under Art 123(2) would also have been possible.
I would summarise G 1/25 as follows
Any statement left in the description
(i) which raises doubts about the matter for which protection is sought or
(ii) which raises doubts about the embodiments which are actually patentable
represent an inconsistency between the claims and the description which has to be removed.
In those two situations it is manifest that
– the matter for which protection is sought is “unclear” in the meaning of Reasons 19 and 35
– doubts about the embodiments which remain patentable arise. Those embodiments are therefore “unclear” in the meaning of Reasons 19 and 35.
In those two situations it is also manifest that there is
– a “real doubt as to the meaning of the claim” and
– a “real doubt” whether a technical teaching, examples, or embodiments do, or do not fall within the claimed subject matter.
Art 84 applies both ways
All those alleging that Art 84 applies only to the claims and not to the description have been proved wrong by G 1/25. This point has never been in doubt and forms part of the basic training of every substantive examiner and EQE candidate.
Reasons 38:
Delete"If, for example, a claim has successfully been amended to meet the requirement of non-obviousness in Article 56 EPC, but a statement in the description, including any drawings, expresses a technical teaching reflecting the claim before this amendment, and this statement thus conflicts with the fulfilment of the requirement of non-obviousness, this inconsistency in the description, including in the drawings, has to be removed".
Reasons 39:
"This is because the requirements of Article 52 to 57 EPC are to be assessed on the basis of the claims as so interpreted. Where the inconsistency materially affects that interpretative exercise, the claim may fail to meet those requirements".
Doesn't r. 39 confirm that r. 38 is only talking about inconsistencies which give rise to real doubt about the meaning of the claims? We can debate whether r. 35 sets a different standard, but there are certainly reasons to believe that it does not.
@ Anonymous-09.09.2026 at 10.34.00 GMT+1
DeleteYou may believe what you want to believe, and dissert about what could be a real doubt when it comes to assess the patentability of a claim. It will not made any change in my reading of G 1/25.
A prime example of a real doubt is to found in the Agfa/Gussi decision of the CFI UPC LD Hamburg. I will not repeat it here.
If you do not acknowledge that in this case, when leaving chromatic in the description there is no real doubt, then, in your views, there will ever be any real doubts.
With respect, it is hard to see how an obiter comment in a first instance UPC decision might provide any insight into how a later EBA decision should be understood and interpreted. For a start, the UPC decision is not cited in G 1/25 and there is no indication in G 1/25 that the EBA is in any way endorsing the view expressed by the Hamburg LD.
DeleteWhilst I do not expect you to change your conclusions about G 1/25, it does not seem unreasonable to expect a straight answer to the question of whether r. 38 relates to real doubts about claim scope. A simple "yes" or "no" will suffice.
@ Anonymous-09.09.2026 at 12.40.00 GMT+1
DeleteI will not dispute that an obiter dictum from a UPC LD carries no binding authority on the EBA. The UPC LD nevertheless came to the conclusion that “chromatic” should have been deleted before grant. It is a direct criticism of the grant procedure and in this respect it cannot be ignored.
This obiter illustrates very well that G 1/25 was necessary. I only cited the UPC LD Hamburg decision to give an example that in this case, there is a real doubt about what is actually covered by the claim. The real doubt stems from the mere fact that Agfa did claim for infringement for “chromatic” in spite of the claim having been limited to “achromatic”.
This example shows very well that the description needs to be adapted. In any case, there should not be given any opportunity to a proprietor to leave in the description any statement which allows a different interpretation of his claim depending on the circumstances.
In reason 38, I fail to see any reference to doubts, whether real or not, or to the scope of the claim. I see therefore no necessity whatsoever to answer a question which not even touched upon in said reason 38.
I certainly do not need to jump over every hurdle you put in my way..
What you do not see in r. 38 is there in black and white in r. 39. You do not have to "jump over the hurdle" of addressing inconvenient facts (the EBA's reasoning in r. 39). However, there are consequences to refusing to address facts which appear to directly contradict your interpretation of G1/25. Those consequences are that everyone else is provided with no good reason to accept your interpretation and plenty of good reasons to reject it.
Delete@ Anonymous-10.09.2026 at 12.18.00 GMT+1
DeleteNeither in Reason 38, nor in Reason 39, I can see any reference to doubts, whether real or not, nor to the scope of the claim. My position is thus still the same.
The subject-matter of a claim is either novel or not, inventive or not, but it cannot be considered half novel or half inventive.
In my blog, comments are not vetted before publication. I do however have the possibility to delete comments I consider not contributing to the discussion, but merely repeating always the same argument. Therefore I have deleted the comment from Anon-08.09.2026
As your are anonymous I do not know if your are the same person commenting in my blog and in this blog.
In the absence of any reaction from Anon. or Anonymous, I can only conclude that for this person, there is no need to adapt the description in the famous Agfa/Gucci case. I am simply of the opinion that, in this situation, G 1/25 requires adaptation of the description.
I am open to any compelling counter argument.
I refer you to the second sentence of r. 39:
Delete"Where the inconsistency materially affects that interpretative exercise, the claim may fail to meet those requirements".
This clearly and unambiguously ties the "inconsistency" (and any objections that may arise therefrom) to the interpretative exercise, namely to the meaning of the claim. The type of Article 56 EPC objection mentioned in r. 38 therefore only (potentially) arises if there are real doubts about the meaning of the claim.
Also, the first sentence of r. 39 starts with "This is because": so there can be no doubt that the explanation in r. 39 relates to the standard described in r. 38.
So remind me why, despite the above facts, you do not agree with the view that r. 38 is only talking about inconsistencies which give rise to real doubt about the meaning of the claims?
I am interested by DXThomas's continued references to the Agfa/Gucci case, in particular his assertions that this case means that certain adaptation of the description is required.
DeleteActually, the opposite is true.
I think that it can be agreed that in that case, the claim was restricted to "achromatic", the description was not amended to delete reference to "chromatic", and the patentee alleged infringement on the basis of a "chromatic" embodiment. The Court commented that "chromatic" should not be relied on and should have been deleted.
But really, did "chromatic" need to be deleted for the claim to be clear and supported by the description? The Court had no trouble determining that "achromatic" in the claims should be interpreted as meaning... "achromatic". Nor did they have doubts that "achromatic" should not be extended to cover "chromatic". The presence of "chromatic" in the description therefore did not cause any lack of clarity or insurmountable difficulty with claim interpretation.
Now you can criticise the patentee for wrongfully bringing an infringement action. That the patentee was confident in alleging that "achromatic" meant "chromatic", or that "chromatic" is an equivalent, is unfortunate. But the efficient way to disincentivise overassertiveness is via an award of costs. Bring a bad case, pay twice.
It is not even clear that a counter-factual case in which "chromatic" had been deleted would run much differently. Agfa still could have alleged that "chromatic" was an equivalent to "achromatic" and the Court would still have been required to decide that issue, presumably with the same outcome.
Requiring complete concordance between the description and the claims or amended in the manner now required by the EPO, involving complex and risky amendments, is not necessary nor is it a welcome feature of European patent practice.
I've always found it hard to believe the patentee brought the infringement action because they thought "achromatic" meant "chromatic". I haven't read anything other than the agfa/ gucci decision in its case history ( so someone correct me if I'm wildly off the mark), but I instead speculate that the following text from the decision is revealing as to the bringing of the action:
Delete"As a starting point it has to be interpreted whether the term “achromatic” refers to the pigment or the base coat as a whole, in other words, if it is sufficient that the base coat contains an achromatic pigment different from black or if it does require the base coat as a whole to have an achromatic colour different from black. The Court construes this feature in the sense of the latter meaning and the parties have finally agreed with this interpretation"
I can more easily believe that the patentee/representative thought they would be on the right side of what is effectively a comprising/consisting of distinction for the make up of the base coat, and when they weren't, felt they had to go on with more tenuous arguments.
If my speculation is correct, agfa/gucci offers less evidence of a need to disincentivise overassertiveness based on inconsistent definitions between the claims and description
Anonymous1: Nail. On. Head.
DeleteIf chromatic had been completely erased from the description, Agfa would still have tried to enforce their patent against Gucci. Even if chromatic had been disclaimed as “not the invention”, Agfa would have argued it was an equivalent achieving substantially the same result on substantially the same way. The EPO needs to stop thinking it can control what happens downstream. It simply cannot. 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.
The G 1/25 test clearly means the description in Agfa would not need to be amended because it is clear what the literal meaning of the claims is. If the claim says achromatic and the description says chromatic, there is no doubt what the claim covers. The description contains unclaimed technical teachings which do not affect the interpretative exercise or cast doubt on patentability. The threshold for amendment of G 1/25 is clearly not met.
@ Anonymous-11.09.2026 at 10.22.00 GMT+1
DeleteI have read Reasons 38 and 39 many times. My conclusion is clear.
If the description allows to give a different interpretation of the matter for which protection is sought when it comes to compliance with Art 56, or all the other Articles mentioned in the order, there is manifestly a real doubt about the meaning of the claim.
I have never said anything different. In a new post on my blog, I have given examples of situations in which there are real doubts about the definition of the matter for which protection is sought when comparing description and claims. The three decisions T 2766/17, T 0758/13 and T 2178/17 quoted therein are perfectly in line with G 1/25.
In T 2178/17, the board decided that the adaptation of the description during oral proceedings before the board is not an amendment in the meaning of Art 13(2) RPBA.
What are you actually after? My mother tongue is not English, but the use of “may” does not render the sentence so vague that it allows to conclude to the absence of an inconsistency.
Whether you like it or not, the adaptation of the description has not gone down the drain. What is at stake is not formal concordance, but substantial concordance. Wait and see until we have some decisions of the boards in this matter.
@ Anonymous-11.09.2026 at 10.22.00 GMT+1
Delete@ Anonymous1-11.09.2026 at 11.53.00 GMT+1
@ Prowsie-11.09.2025 at 15.56.00 GMT+1
@ Anon-11.09.2026 at 15.56.00 GMT+1
It will not come as a surprise that I find your reasons why in the Agfa/Gucci case the description need not to be adapted anything but compelling.
Whether Agfa would have attacked even if “chromatic” would have been deleted, is pure speculation. When you read the decision carefully, Agfa brought the infringement action because "chromatic" could also fall under "achromatic". Taking the content of the description at face value, the court held that, according the definition of “achromatic” in the description itself, “chromatic” was not what was defined in the claim. In other words, the court can be said to have applied G 1/24 beforehand.
We have to keep to the facts. Fact is that Agfa attacked on the basis of “chromatic” since this term was not deleted. On top of this, the court complained about the non-deletion of the term.
It is manifest that the court was not fooled by Agfa’s interpretation of the claim. This is however not a reason to claim that “chromatic” should have been left in the description.
As the description has not been amended, it contained statements which were at odds, to be clear, the opposite, to the definition of the matter for which protection was sought in the claim.
If this does not imply reasonable doubts, then no adaptation of the description will ever be required. This is what you all want, but I do not take this from G 1/25.
The G 1/25 test clearly requires that the description in Agfa has to be amended. The claim limited to “achromatic” is without any doubt, clear as such, but not supported by the description which says something quite different. This is a matter of fact and not of opinion.
It is particularly revealing to read: “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.
I would say that G 1/24 and G 1/25 are the best weapons against Angora cats or against changing interpretation by the proprietor depending on the circumstances.
That you do not like this perspective, is your problem, not mine.
Moreover you seem to forget that in one situation you will be patentee and in another one, you will be opponent or potential infringer. The EPO has the duty to grant solid patents and not patents which can be interpreted at the whim of the proprietor.
Part 2
ReplyDeleteA positive change in practice
In my opinion adaptation of the description at the R 71(3) stage should come to an end, and it is rightly so.
It is to be welcomed that the EBA has suggested that the boards align the description with the claims before referring the case back to the first instance for grant or maintenance in amended form. I recall lengthy discussions – ultimately fruitless and exhausting – between the proprietor and the opponent regarding the adaptation of the description.
Some boards will try to sidestep this, as it is merely an obiter dictum and not part of the decision. Let us hope that the majority of boards will comply with this suggestion.
One question nevertheless arises: at what stage must an amended description be submitted by the applicant/proprietor? At the time of filing auxiliary claims in first instance, when lodging an appeal or in response to an appeal, or during the oral proceedings before a board? Some boards are rather restrictive in their understanding of the three levels of convergence set in the valid RPBA. Better be safe than sorry.
Inconsistency after limitation of the claim
When for instance, the original claim consists of A and B, and in view of the prior art, the claim is later limited to either A or B, it is not possible to allege that claim A is supported by B left in the description or claim B is supported by A left in the description. This is exactly what Agfa did with chromatic/achromatic and was queried by the UPC LD Hamburg. In such a binary situation, either A or B has to disappear from the description.
It can well happen that some embodiments are not any longer falling under the claims. Those do not have to be deleted, but they cannot be held to be supporting a limited claim and have to be marked accordingly as they manifestly contradict the claimed inventive teaching.
G 1/24 and T 439/22-2
As regards T 439/22, ‘reading into the claim’ the interpretation found in the description does not simply mean “virtually” or “intellectually” reading the interpretation found in the description into the description, but actually replacing the definition in the claim with that found in the description.
Where there are two different definitions of the matter for which protection is sought in the description and in the claim, it is not possible to conclude that the description and the claims are consistent with one another.
Leaving the description and the claim as they stand does not resolve the manifest inconsistency. By applying the principles of G 1/24, see Reason 18, the inconsistency is resolved by replacing one definition by the other, at the discretion of the applicant/ proprietor.
In any event, in light of G 1/24 and G 1/25, it is clear that a patent cannot be its own dictionary, as some seem to conclude from T 439/22.
Any amendment of the original disclosure and the original claims has to be compliant with the requirements of Art 123(2). It is for the applicant/proprietor to keep an eye on this point and the EPO to check this compliance.
In the comments on my blog, I gave the example of the famous Agfa/Gucci case when it comes to delete subject-matter in the description which is clearly at odds with the limited claim.
I also gave the example of the famous pemetrexed case. In my opinion, the statement left in the description was manifestly infringing Art 123(2).
This is why I have always considered that G 1/25 would be the continuation of G 1/24. Applying G 1/24 and writing into the claim the definition in the description, or vice-versa, represents a first step in resolving any inconsistency.
When readers address your 2-part posting above, Daniel, I should like them also to address the hypothetical we two have looked at in your own blog. It (somewhat polished in the meantime) goes like this:
DeleteI file at the EPO with a process claim 1 that recites citrus. The description describes four experiments, each using a different citrus fruit (orange, lemon, lime, grapefruit). I am careful in my (US-style) drafting to avoid the word invention, using instead the word "disclosure" so that none of my disclosed experiments is said to be an "Example" or "embodiment" of "the invention". They are all simply disclosures in support of the subject matter for which protection is sought, to lift the disclosure up to a level sufficient to satisfy Art 83 EPC.
But then it turns out that "citrus" is untenable. There is an accidental loss of novelty of "orange". So I amend the claim to "lemon" and file a divisional to each of "lime" and "grapefruit".
I see no inconsistency between any of my narrowed claims and its supporting description, and no failure of clarity in any of the narrowed claims, and so see no need to amend the as filed description.
If I understand you correctly, however, Daniel, you as an EPO Examiner would invoke G1/25 and follow the UPC in Hamburg, would vigorously disagree, and would insist on amendment to the description, because an "inconsistency" is present, one which creates real doubt about what the claim covers. Readers, what do you think? Is there an "inconsistency" here? Or not?
Incidentally, on the subject of improving legal certainty for judges of what the claim "covers", another citrus fruit is the so-called Chinese grapefruit, the pomelo. Does amendment of the description to delete orange, lemon and lime help or hinder the court's deliberations whether the grapefruit claim, under Art 69 EPC, "covers" the pomelo? If I were a judge, I think I would prefer to have the description as filed as the basis of my deliberations
@ Max Drei,
Delete3 comments
1) On September 5th you wrote in a comment on my blog:
The trouble with hypotheticals is that they over-simplify and so hinder rather than help debate. What are doing here? You come with a further hypothetical.
It is always possible to construct a hypothetical in such a way that it favours your point of you. It reminds me of a former LQM of the boards. Otto Bossung was well known to always come with hypotheticals, to the effect that in the end, nobody any longer listened to him.
2) Accidental anticipation
In your hypothetical you speak expressis verbis that there is an “accidental loss of novelty” of "orange".
In case of an accidental anticipation of “orange”, G 1/03 applies. When G 1/03 applies, the only amendment needed is to acknowledge the source of the accidental anticipation in the description and to add a corresponding disclaimer in the claim. No further change is needed.
In this situation it would be stupid for an EPO Examiner to invoke G1/25. Your hypothetical has absolutely nothing to do with the UPC case in Hamburg.
In the UPC case in Hamburg, there were two equivalent alternatives in the originally filed claim. One of those was novelty stricken, but not accidentally. Therefore, I cannot follow in the slightest your theory and your comparison.
By the way, I recently checked the number of cases in which an accidental was acknowledged by the boards. It is not more than an handful of cases, although applicants/proprietors have tried a lot of times to come up with an accidental anticipation.
3) Art 69 and the pomelo
For a start, it not within my competence to decide what a court would or should do when applying Art 69. The EBA has further made clear in G 1/24 that Art 69 applies primary in post-grant and post-opposition possible.
On top of it, your hypothetical is so construed that there is no reason to amend anything but acknowledge the accidental anticipation.
A further comment about US style of drafting
DeleteWhether you are careful and draft in US-style avoiding the word invention, using instead the word "disclosure", it will not change the fact that in European practice that the claim define the matter for which protection is sought. This will be either a generalisation of a series of embodiments, or even from a single one.
It is therefore an illusion to conclude then that none of your disclosed experiments is said to be an "Example" or "embodiment" of "the invention", and that it will impress the deciding bodies of the EPO.
I have also seen US originating applications in which each sentence starts with “May”. This does also not work at the EPO. If everything is optional, nothing is optional, cf. T 1203/13.
Whatever US applicants and their European representative might think or argue is irrelevant when an application ends in Europe. The EPC applies and nothing else.
Daniel, I must straightaway say how much I admire the consummate ease with which you dismiss and wipe away my hypothetical. With the benefit of hindsight, I wish I had not included the word "accidental". I should have followed my own earlier advice and rigorously eschewed any use of any hypothetical.
DeleteThanks too for your riff on what happens when US drafting style collides with an EPO tribunal. Yes, when the claim expresses a technical feature as essential and the description declares it to be merely optional, we have an inconsistency. We agree on that, at least. My hypothetical was trying to explore elsewhere, the boundary between an inconsistency and the absence of an inconsistency. I think that we need to know whether or not an inconsistency exists, before we go on to enquire whether it is one which causes a non-compliance with the EPC. You might think my question unnecessary, bothersome and trivial but I shall continue to wonder about it until I get an answer that convinces me. That US judge who said of obscenity that "I know it when I see it" would presumably say the same of an alleged "inconsistency". I think we can do better than that, without resorting to circular logic that, if using a holistic approach, I am still not clear what a claim means, that is all I need, to prove that the cause of the lack of clarity is the presence of an "inconsistency".
Max Drei,
DeleteI must confess that I was surprised when you spoke about an accidental anticipation. It did indeed made it very easy for me to dismiss your argumentation.
I think that with G 1/25 we know the meaning of the word inconsistency and when it comes into play: see Reasons 23. In reasons 33-35, the EBA explained what an inconsistency, with respect to Art 84 means. In reasons 36-39, the EBA explained what an inconsistency, with respect to Art 52-57 means. In Reasons 40-42 the EBA did not say much about inconsistencies and further Articles of the EPC. What applied to Art 84 and 52-57 would apply mutatis mutandis.
This allowed the EBA to refer in its order to Art 52 to 57, 76(1), 83, 84, 123(2) or 123(3).
I do not think that your question is unnecessary, bothersome and trivial, but I doubt you will get more out of the EBA. We will now have to see how the boards apply G 1/25. I personally do not think that much will change, besides some excessive requests from examiners in matter of adaptation. I call excessive request any request which is not properly reasoned as to why there is an inconsistency affecting Art 52 to 57, 76(1), 83, 84, 123(2) or 123(3).
OK Daniel. That helps. But I remain pessimistic. I think it will be trivially simple for an EPO Examiner, following G1/25, to find reasoning, if they are minded to, why there is a need for deletion or amendment of a string of words in the description that they don't like.
DeletePerhaps ever-greater pressure from EPO management on Examiners, to attain ever-higher levels of productivity, will be the most effective deterrent to the raising of ever more objections based on the Examiner's perception that an "inconsistency" is "leading to" a failure of the application to comply with the EPC.
@ Max Drei-09.09.2026 at 12.41.00 GMT+1
DeleteIf the only hope that G 1/25 is not applied with too much rigour in view of the production pressure of the higher management, then the situation is rather bleak.
I would in any case prefer to see EDs doing better searches and examinations rather than fiddling with the description, when one sees what happens with the patents being opposed.
When looking at board decisions published up to this day, in 2026 only 11,7% of opposed patents survive an opposition. 45,7% of opposed patents are revoked and 33,3% of opposed patents are maintained in amended form, i..e limited in scope. 9,3% of cases are remitted to the OD.
In only around 8% of the cases, opponents have cited prior art which could genuinely not be found in the EPO search documentation. When the opponent comes with prior Art under Art 54(3), this does not look good.
Around 70% of the decisions of opposition divisions are set aside by the boards. This does not look good either.
Those figures have not changed over the years.
Those figures worry much more. What does it bring to adapt the description at grant, when in opposition patents are, to a large extent, coming out maimed.
Here Max Drei. If the controllers of quality inside the EPO find it easier to police compliance with Art 84 than whether the search was thorough, Examiners will presumably continue to do what helps their career progress
DeletePart 3
ReplyDeleteThere is a further need for adapting the description:
If the claim needs to be limited in relation to prior art that is closer than that available to the applicant at the time of filing the application, the applicant must replace the original prior art. This also constitutes a form of amendment to the description. See Guidelines H-V, 2.4 and F-II, 4.3.
There are two options available:
1) In case of a two-part claim, a simple reference to the preamble of the claim is sufficient. That US applicants do not like this form is not a secret (Jepson claims)
2) In case of a single-part claim, all the features disclosed in the prior art document that led to the limitation of the claim must be set out.
There is no way out when it comes to acknowledging the “proper/better” prior art. It is left at the discretion of the applicant/proprietor, but it is mandatory.
An applicant/proprietor might, for whatever reason, refuse to acknowledge the “proper/better” or closer prior art in examination or in opposition, but then, you should not be surprised that your application is refused or the patent is revoked.
Conclusion
I am therefore not sharing the view that G 1/25 does not bring any substantive gain in legal certainty. To me the contrary is true. That it means more work for attorneys, greater costs for clients, is however a direct consequence of G 1/25.
The EPO is meant to deliver a patent valid in lots of countries, and the EBA has found a way to insure that the applicant/proprietor cannot vary the interpretation of the claims as he thinks fit after grant or opposition. To me this is the net result of G 1/24 and G 1/25.
It might take a while until applications are filed which take into account G 1/24 and G 1/25, but I see in those two decisions a clear benefit for third parties. They are not any longer left in the dark as far as interpretation of the claims is concerned.
This will not withhold national or supranational jurisdictions to decide infringement or nullity of a patent as they think fit. At least the proprietor will not be allowed to vary in its interpretation of the claims.