EP 3 161 066 B1 relates to polyethylene compositions used for injection moulding.
Brief outline of the case
The opposition was rejected and the opponent appealed.
The reason for the rejection was that two resin compositions DMDA-8904 NT7 and DMDA-8907 NT7 were considered not belonging to the prior art under Art 54(2).
The board considered that, in view of G 1/23, the two compositions belonged to the prior art under Art 54(2). It followed that claim 1 as granted was lacking N over the two compositions. The amendments to claim 1 of AR I did not overcome the objection.
The patent was eventually maintained according to ARII which was deemed inventive over the two resins taken as CPA or over D6=US 7,300,988, not mentioned in the ISR established by the EPO.
The file was remitted to the OD for the adaptation of the description.
We will concentrate on AR II and the two resin compositions used as CPA.
The proprietor’s point of view on the resins as CPA
For the proprietor the knowledge of the method used to obtain the DMDA resins is essential in order to achieve a resin composition as defined in operative claim 1.
The proprietor argued that the two commercial polymer grades, DMDA-8904 and DMDA-8907, could not be reproduced by the skilled person without undue burden, due to the lack of crucial information about the catalysts and process conditions used in their manufacture.
Reverse engineering these products would not allow the skilled person to reconstruct the precise manufacturing process, as industrial polymer grades depended on complex and proprietary manufacturing methods involving specific interactions between a catalyst, process parameters, and polymerisation unit.
The proprietor further argued that the skilled person would have no motivation to modify a commercial product since the resulting product would require (re)certification.
Under these circumstances, it was not obvious for a skilled person starting from the DMDA resins as the closest prior art to obtain a composition according to claim 1 of auxiliary request II.
The opponent’s point of view on AR II
During the OP, the opponent argued that the DMDA resins were used in the same technical field as the claimed invention, making them a realistic starting point for further development.
Furthermore, the ability of the skilled person to reproduce these resins was irrelevant to the choice of the CPAt. The decisive point was whether the skilled person had access to them or could produce a resin with similar properties.
The opponent took the view that the skilled person would know how to adjust the density of the DMDA resins while maintaining the other properties, MIE, MIF and ER values, within the ranges defined in claim 1.
The board’s decision on AR II
The central point of dispute between the parties was whether a non-reproducible product can be selected as the CPA for assessing IS. In that regard, it will be assumed, for the benefit of the party arguing that the DMDA resins cannot be selected as the CPAt, that these resins could not be reproduced by a skilled person before the effective filing date of the opposed patent.
For the board, it was not in dispute that a PPU may be selected as the CPA for the assessment of IS.
This finding is supported by the case law, such as T 1464/05, Reasons 5.2.2, T 0660/16, Reasons 7.2 or T 2463/22, Reasons 6.4 to 6.6.3.
The question is, however, whether this also applies to commercial products which cannot be reproduced by the skilled person.
For the board, G 1/23 does not exclude that a non-reproducible product be selected as the CPA. Instead the specific circumstances of the case must be analysed.
It is known that the resins DMDA-8904 NT 7 and DMDA-8907 NT 7 were polyethylene compositions which could be used for injection moulding and were characterised by good impact stress crack resistance and processability.
These characteristics correspond to the technical field and the alleged advantages of the compositions claimed in the opposed patent. In the board’s view, this information makes commercial resins DMDA-8904 NT 7 and DMDA-8907 NT 7 immediately relevant for a skilled person working in the present technical field and wishing to provide alternative or even improved compositions.
Even assuming that the skilled person would not know how to reproduce these resins, the board did not consider this point to be a sufficient reason to disregard them as CPA.
First, the DMDA resins being commercially available, the skilled person could use them as such or modify these resins, within the limits of its competence, to obtain different properties. This situation essentially corresponds to the example given in G 1/23, Reasons 96.
For the board, the skilled person could have analysed the internal composition of the DMDA resins and, within certain limits, attempted to reproduce some of their properties.
In any case, there is no apparent reason why the skilled person would have considered a priori, i.e. before identifying the differences and establishing the objective technical problem solved, that any further improvement necessarily required a modification of the specific synthesis conditions employed for the preparation of the commercial product, which were not known, and excluded the product as a possible starting point on that basis.
Accordingly, even if the DMDA-8904 NT 7 and DMDA-8907 NT 7 resins were not fully reproducible, they are nevertheless a realistic starting point for further developments in the present technical field.
Contrary to the proprietor, the board did not regard the absence of the knowledge of the method used to obtain the DMDA resins as a valid criterion for not selecting them as CPA.
In this respect, the extent to which a non-reproducible product must be modified to obtain the claimed subject-matter, and the level of knowledge about that product and its manufacture required to achieve it, are not considered relevant for determining whether the product can be regarded as the CPA.
These aspects instead relate to later stages of the problem-solution approach, namely identifying the distinguishing features between the product and the claimed subject-matter and/or assessing whether the skilled person, starting from that product, would have been able to obtain the claimed one.
Under these circumstances, the board considered that commercial resins DMDA-8904 NT 7 and DMDA-8907 NT 7 can be selected as starting point for assessing IS.
For the board, the problem to be solved was, at best, the provision of an alternative composition.
Even if the selection of parameters defined in claim 1, density, MIE, MIF and ER values, is considered arbitrary, it would still be necessary to establish that the preparation of a polyethylene composition meeting those parameters was itself obvious.
The decisive issue is thus whether the skilled person, relying on the information available in the prior art and, if necessary, on CGK, would have been able to identify the measures leading to a composition that fulfils the parametric definition of claim 1, without resorting to undue experimental effort.
The exact processes by which the DMDA resins are prepared are not disclosed in the prior art and are likely to constitute confidential know-how of the manufacturer, i.e. the opponent.
Although the opponent explained in general terms which practical steps could be followed – such as the use of a Ziegler-Natta catalyst in a gas-phase fluidised-bed reactor as part of the UNIPOLTM process – it has not been shown that, by following these indications and with only routine experimentation, the skilled person would have been able to reproduce the DMDA resins or obtain materials having comparable properties.
The board accepted that a skilled person could indeed increase the density of a polyethylene, the evidence on file does not allow to conclude that a skilled person would have no difficulty in maintaining the other characteristics of the polymer. This is all the more problematic given that the ER value of DMDA resins, 0.50 for DMDA-8904 NT 7 and 0.49 for DMDA-8907 NT 7, is relatively close to the upper limit of the range defined in claim 1, 0.40 to 0.52. The opponent did not provide evidence of what a skilled person would have concretely done to adjust this parameter.
For these reasons, whether the skilled person would be able to identify the measures which would, starting from the DMDA resins, lead to a polyethylene composition that meets the parametric definition of claim 1 without having to resort to an undue amount of experimental work remains a matter of conjecture.
The board concluded that the invention defined in claim 1 of AR II has not been shown to be obvious to a person skilled in the art starting from any of the DMDA resins as the CPA.
Comments
Whilst accepting that a commercial product can be used as CPA, the board went down a different route: the obviousness hinged on whether the skilled person would have been able to reproduce the DMDA resins or obtain materials having comparable properties. This was not sufficiently demonstrated by the opponent and hence the lack of IS was not demonstrated.
The conclusion is understandable, but claiming that a commercial product is prior art under Art 54(2), and can even serve as CPA, but acknowledging IS of a claim due to the fact that the commercial product is not reproducible, is somehow contradictory and shows some problems in the reasoning of the EBA in G 1/23, at least when it comes to assessing IS.
In T 438/19, commented in the present blog, the patent leading to G 1/23 was revoked as the proprietor withdrew the approval of all requests. In the board’s communication under Art 15(1), the board argued that the product ENGAGE could be replaced by a similar product with the help of the teaching of a further document. This could be a practical route.
In T 143/24, commented in the present blog, the board held that products sold under a confidentiality agreement are not part of the prior art. Reasonable doubts are permitted whether this decision is in conformity with G 1/23, as selling products under a NDA allows to easily circumvent G 1/23.
In T 807/23, commented in the present blog, the board has apparently applied G 1/23, but sent the case back to the OD without taking any decision on the substance.
We will have to wait for more decisions to see in how far G 1/23 is actually applicable when it comes to assessing IS.
Comments
53 replies on “T 1044/23-On the use of publicly available products as CPA under G 1/23”
Haven’t yet read the decision but first impression is positive. Why: because I want to defend as real world and common sense the EPO approach to obviousness to sceptical engineers and business managers. This decision helps me in that, doesn’t it?
Dear Max Drei,
The decision is positive as it brings to light the problems created by G 1/23.
On the one hand, commercial products are CPA, but on the other hand the lack of reproducibility does not allow to decide that that the claimed composition is obvious. This is a catch 22 situation.
This is actually the fundamental problem with G1/23. The example with the Cola in Reasons 96 of G 1/23 is oversimplistic and not helpful.
Simply adding that” the relevant technical teaching that a skilled person would take from such a product is always case specific – it depends on both the product in question and the invention under examination”.
Decisions of the EBA are to help coming to a reasonable interpretation of the EPC and to give some guidance to the EPO and its users. Reasons 96 is not helpful at all.
There is a long line of case law according to which a non-reproducible piece of prior art cannot be taken against a claim, be it for novelty and/or inventive step. Is this case law not any longer valid, or in case of a commercial product, is this line of case law is to be ignored? Difficult to follow either positions.
G 1/23 will be the cause a lot of headaches an/or difficulties.
As to Catch 22, Daniel, I am not yet persuaded that we have one here. Catch 22 comes from a work of fiction but from real life comes the Art 123 2/3 fatal trap which for me is fiction become fact. That is a truly Catch 22 set up, but not here I think.
In real life the most realistic starting point can be a market leading product which is protected by a trade secret. Who can deny that? The Boards must find a way to cope with that reality. The optimal way forward is the one that reconciles legal certainty with compatibility with the real world, no matter how challenging that is, for the ED’s, in any given messy matrix of fact. This decision gives the ED a roadmap. Or not?
Dear Max Drei,
I will not argue whether we have a catch 22 situation or not.
Another way to describe the situation is the following: the cat bites its own tail (Original in German: die Katze beißt sich in den Schwanz).
The situation is simple: a commercial product can be used as CPA, but if it is not reproducible, it cannot help in deciding whether the claimed subject-matter is inventive or not, and then we are back to square 1. The present decision boils down to a vicious circle and shows that reality has not been thought through in G 1/23.
The present decision is therefore not helpful for EDs and ODs.
One way which appears practicable is the one the board considered in its the communication under Art 15(1) RPBA in T 438/19.
The board argued that the product ENGAGE could be replaced by a similar product with the help of the teaching of a further document, and that this would lead to a lack of inventive step. This could be a practical route.
Since the proprietor has (wisely?) withdrawn all requests, we will not know the last word on this point.
Sorry, Daniel. I am not following you. You suggest that “a practical route” would be to start from the “ENGAGE” product, use a further document to teach how to make “a similar product” and then continue on from that similar product.
Is that not still using Engage as the “starting point” for the patentability analysis under Art 56 EPC?
In my view, the issue of whether a non-reproducible product commercially available before the filing or priority date is, or not, an admissible closest prior art for the assessment of inventive step should take into account the exclusionary effect of the patent supposedly granted, assuming that the commercial product is denied as a CPA and the EPO finds inventive step. This is not to mean that validity should be assessed solely on the basis of such an analysis, but simply that the implications of the validity decision in such a situation deserve close attention to avoid absurd or awkward outcomes.
My question is then : what is the foreseeable impact on the exclusionary effect of the patent if a novelty-defeating product was put on the market before the priority date ? At first glance, the scope of the exclusionary effect is seriously impaired, since the commercial availability of the product creates a “safe harbour” for the activities involving the product – not just the sale or lease of the product, but also any use of the product by its buyers/possessors, and any use of the information derivable from the product as well. In other words, an unfettered freedom to use for the buyers. A big fish falling through the net of the patent ! Does it still make sense to grant a patent when this is acknowledged ?
Mr Hagel,
I think that we should not confuse the cause and the effect.
The cause is the commercially product available before the filing or priority date.
The effect is whether it can be used to attack N or IS.
If a non-reproducible product is commercially available before the filing or priority date, it cannot be said that it is N destroying , when its composition and/or its method of manufacture it is not known. This corresponds to a long line of existing case law: a non-enabled product cannot be opposed to a patent. It might well be used theoretically as CPA, but if it turns out that it is non-enabled, its use as CPA leads to nowhere. This was the case of the AR in dispute.
In the conference on case law 2025 of yesterday, when presenting G 1/23, Mr Bokor was of the opinion that a product sold under a confidentiality agreement is not publicly available. When a commercially available product is not enabled as the producer keeps its composition and method of manufacture confidential, this boils down to a commercialisation with an implicit confidentiality agreement.
If the properties made publicly available correspond to or are similar to those of the claimed properties, then the situation is different.
If a product identical to the non-enabled commercially available product can be obtained with the help of the available prior art, or if the known properties of the commercially available product correspond to those claimed, then N can be challenged.
If a similar product to the non-enabled product can be obtained with the help of the available prior art, then IS can be challenged. The fact that the commercially available product is not enabled becomes then irrelevant since it can be replaced. This is what I call the practicable route, as explained by the board in T 438/19.
It still makes sense to grant a patent if the only prior art available, a commercial product marketed before the effective date of the claim, is not enabled or the information about its properties do not correspond to what is claimed.
This was the situation in the present case. Claim 1 as granted was N stricken as the properties made publicly available corresponded to those claimed.
When it came to the AR, the non-enablement was manifest, and hence the claim could not be considered obvious.
The non-enabled product could not be replaced by a similar product.
This was not the case in T 438/19. The subject-matter of the claim was considered obvious as the non-enabled product could be replaced by a similar product.
Dear Max Drei,
In matters of practicable route, I would like to explain myself.
The only difference with claim 1 as granted and example 3 of D1 using ENGAGE®8400 is the aluminium content. This content is not known for ENGAGE®8400 and cannot be determined in the absence on any indication of the synthesis method. This is why the proprietor considered example 3 of D1 as non-enabled.
In the communication under Art 15(1) RPBA, the reasoning did not start with ENGAGE®8400 as CPA, but with example 3 of D1, and by replacing ENGAGE®8400 by a similar material.
For the board, it appeared that the skilled, person would implement the teaching of Example 24 of D18 in order to obtain a copolymer similar to ENGAGE® 8400, as he would be guided by the general teaching of D18 and therewith come to the aluminium content of claim 1.
This is what I consider a practicable route: replace the non-enabled commercial product by a product with similar properties. It is not to use the commercial product as such as CPA, but to start with a CPA comprising the non-enabled commercial product ,and replace the latter. This might not always be possible, but in T 438/19 it led to the proprietor to withdraw all his requests and the approval of the patent as granted.
I hope my position is now clearer.
It is clearer now, Daniel, thanks.
Mr Thomas,
The effect you refer to relates to validity. But you missed the point of my comment. I was wondering about the exclusionary effect of the patent, assuming it was granted.
Thank you for mentioning Mr Bokor’s opinion. It does not apply to the commercial products cited in the decision, as they are sold by the opponent Dow Chemical without restriction, as noted in the decision (reason 2.5.5). And the fact that the seller keeps secret the composition and method of manufacture secret is a fact, which cannot be equated to an implicit confidentiality agreement. The buyer is free to reverse engineer the product and it is possible that advanced analysis techniques provide insights as to the composition and method of manufacture of the product. In other words, trade secret protection is at the seller’s own perils.
Getting back to validity and the IS assessment, it seems to me that the assessment of obviousness is substantially weakened when it is based on a technical problem limited to finding an alternative, as such a definition is devoid of technical content. Is there a « technical problem » without technical content ? I would be interested in your views on this.
Mr Hagel,
My contribution was indeed about validity. I dare say something about infringement as I lack any competence in it, and this is the domain of national courts and the UPC. The EPO can only assess validity, but has no influence on how national courts and the UPC assess infringement.
The patent has indeed an exclusionary effect, but the EPO can only refuse the grant or the maintenance of a patent on the basis of the EPC, but not on the fact that the patent might have an exclusionary effect. This is the essence of the patent system.
I agree with you that the seller goes into a risk when keeping for instance the method of manufacture as a trade secret. Reverse engineering is nowadays a good possibility, but it also has limits, see T 438/19. In the absence of knowledge of the synthesis parameters, Engage 8400 was not reproducible, as the exact conditions of synthesis were not known. .
Even if the buyer can reverse engineer the product, what can he do with the information? It all depends on the contract between seller and buyer. This type of contract is rarely made public. At best a bill is made public, but it is then redacted, or a proof of delivery is submitted with possibly an analysis report.
What is public, is only that some goods, sold under a trademark have been delivered at a given time. As decided in T 143/24, if the sale is subject to a NDA, confirmed by Mr Bokor, then the buyer cannot do anything with the information even in case of positive reverse engineering.
This is why I consider the sale of a product under a trademark, which at the same time is not reproducible to be equivalent to a sale under NDA.
As far as considering the OTP as being a mere alternative, I can agree with you that the situation might not look absolutely satisfactory, but what else can be brought instead?
In this respect, I would like to refer T 769/23, also commented in the present blog. In this decision, the board has tried to go away from the standard OTP=an alternative when there is no different effect.
It had never occurred to me that the problem of finding a technically competitive alternative to a freely-marketed and successful product, a product which competitors find themselves unable to reproduce, is not a “technical” problem.
Finding a design alternative to a market-leading fashion product though, is to my mind a different matter. seldom will that be a “technical” problem.
As ever, stating a principle is easy, whereas applying it to any given matrix of fact is often hard. To preserve the high reputation of the EPO, the EBA should try its best to formulate principles which the lower levels can fairly and justly apply to any given set of facts, in such a way that both sides of the dispute cannot legitimately criticise.
If patent law and trade secret law is working as it should, then the prior art “starting point” will often be a successful product that is increasingly taking away market share from worthy competitors. the EBA should recognise that reality and explain to the Boards and Divisions how to cope with that reality. The EBA has made a good start with that task, no?
Dear Max Drei,
In reply to your comment that “the EBA should try its best to formulate principles which the lower levels can fairly and justly apply to any given set of facts, in such a way that both sides of the dispute cannot legitimately criticise”, I will quote the last part of Reasons 96 of G 1/23:
“There are no formal and strict rules as to how a non-reproducible but publicly available product or any of its properties can be taken into account when inventive step is examined. The relevant technical teaching that a skilled person would take from such a product is always case specific – it depends on both the product in question and the invention under examination.”
The EBA might have recognised a given reality, but has NOT explained to the Boards and Divisions how to cope with that reality. The EBA has by no means made a good start with that task.
The comparison with natural products is an interesting point in G 1/23, but natural products can, sooner or later, be reproduced with present and future technologies, but this is not necessarily the case for man made products resulting from the intricate synthesis of chemical products. The composition of a product sold under a trade mark might also change with time, but if it is not known how to obtain this product, it becomes irrelevant when this product has actually come on the market.
To sum it up, G 1/23 is one of the worst decisions of the EBA.
As I have said earlier to you, G 1/23 will be the cause a lot of headaches an/or difficulties.
Daniel, I am sure you are right that G1/23 will bestow on Division and Board members what you call “headaches” and what I would call “painstaking finding of the unique matrix of fact in the case and then correct use of those facts to derive the correct result under EPO-PSA”. The “state of the art” is everything (made available) much more than the content of the EPOs database, and every decision under Art 56 EPC must take account of it. Finding the facts is hard work. Inevitably, each decision will produce “headaches”. Therefore, criticism of G1/23, that it will cause headaches seems to me, Daniel, a little unfair on the EBA.
Another thing. In English law jurisdictions, the strict application of “Binding Precedent” can lead to results that many will find unfair or unjust. For the EBA to give the Divisions and Boards freedom to get to a fair result will, I think, enhance rather than damage the reputation of the EPO.
Dear Max Drei,
If I have at all uttered a criticism of the EBA, is that in in Reasons 96, the EBA has not provided any guidance as how to proceed and leaves it entirely to the ODs and the BAs to look how they can best deal with such situations. This is not what is to be expected from the EBA.
I fully agree with the EBA that commercially available products cannot be ignored, but how to proceed with non-enabled commercial products is missing in the whole decision. Som guidance would have been helpful.
Any decision finding goes implies more than often headaches, and this is normal, but the conclusion of the EBA is somehow too short. In the conference on the case law of the boards, Mr Bokor could not dispel the worries expressed in the questions raised by the audience with regard to G 1/23. G 1/23 might cause more headaches as necessary.
The notion of “Binding Precedent” is not a feature of continental law, so that I will not comment further, but that it is not applicable on this side of The Channel.
Mr Thomas,
I will be less critical. First, I think that as a matter of principle, it is a good thing that in G 1/23, the EBA has strictly applied the EPC, without adding conditions. Article 54 defines prior art as publicly available, this is the case for a commercial product, regardless of whether its composition or method of manufacture is known. There is no reason for adding any condition, especially since article 54(2) defines prior art as « everything made available to the public ».
The significance of the term « everything » was emphasised in T 2101/12 which rebutted the interpretation of Art 54(2) provided in T 172/03 which had excluded from the scope of the prior art non-technical elements. Guideline G VII-2 which had incorporated the restriction of T 172/03 was eventually revised.
For the sake of clarity, it is also justified to have cut the Gordian knot in this way. The case law is now clear. It is particularly important for practitioners that the definition of the prior art be as simple as possible.
I agree with what you say but would add that Article 54 EPC defines the state of the art includes everything made available to the public “by use, or in any other way”. This an even stronger reason for the EBA to take the view that it took. For this reason, I can only agree with the general idea underlying G1/23. It may cause headaches but the wording of the law often does cause headaches.
Mr Hagel,
As explained to Max Drei, I fully agree that commercial products cannot be ignored. What is disappointing is that after this logical conclusion, the EBA has not given any kind of guidance. However this should be the role of the EBA.
I never understood T 172/03. The problem is that it has been cited in many other decisions. In this respect the COMVIK decision made a clear distinction between technical and non-technical features. The latter cannot be ignored, but they cannot contribute to the inventive step of technical items.
The definition of the prior art might be clear, but not the way to deal with such prior art.
Should we conclude that all the long lasting case law on non-.enabling prior art is to be thrown overboard? I do not think so. Otherwise it will be easy to fill descriptions with references to commercial products with just indicating some properties bur knowing well that their disclosure is non-enabling, or to sell commercial products with NDAs so that the public is left in doubt with what is actually contained in those.
A disclosure is non-enabling if it cannot be reproduced by the skilled person without undue burden. This is the case for lots of commercially available products, especially in the field of polymers. They cannot be reverse engineered. This is a reality which has been ignored by the EBA in G 1/23.
@ Anonymous,
I refer to my comments to Max Drei and to Mr Hagel.
Commercially available products are certainly part of the prior art, but this is only half the story.
The other half has been ignored by the EBA. What if this prior art is not enabled, and cannot be reproduced by reverse engineering?
Commercial products, beside food for humans, are rarely sold with a list of ingredients, and even in this case the list is not always complete. Here is the problem, and this problem has been dodged by the EBA.
If the commercial product is not reproducible, then that implies, absent evidence to the contrary, that the skilled person is not in a position to produce an adapted product based on the commercial product, and so there is a tendency for it not being considered part of the prior art for that situation.
If the commercial product can be used as is for incorporation into a new product, such as the use of a commercial product for producing a polymer layer in a solar cell arrangement, then the resin is part of the prior art, again absent evidence to the contrary.
@ Pete Perfect,
If the commercial product is not enabling, or if the skilled person is not in a position to produce an adapted product based on the commercial product, there is not just a mere tendency that the product is not part of the prior art, it’s a matter of fact that the commercial product is effectively not part of the prior art.
I do however have some difficulties in following the logic that a document is first chosen as belonging to the prior art, or even can be used as CPA, but then discarded as not belonging to the prior art as not enabled or as not allowing to produce a similar product.
If the commercial product is enabled, or if a similar product can be produced on the basis of the commercial product, there is no doubt whatsoever that the commercial product belongs to the prior art. This was the position of the board in T 438/19.
There are thus two lines which can be followed
– The one in the present decision, but then the commercial product is not part of the prior art, whatever G 1/23 might say;
– The one suggested in T 438/19, and then the commercial product is part of the prior art, according to G 1/23.
Do you see any other possibility?
I think the situation is covered in section 8.7.3 of the referring board’s preliminary opinion after G1/23 (https://register.epo.org/application?documentId=MG9GJIBAGEQQQ3U&number=EP11830390&lng=en&npl=false ), and here I would follow the logic of the patent proprietor. If the commercial product needs to be modified but the skilled person is not in a position to manufacture the commercial product, it follows that the skilled person cannot produce a modified product.
If the commercial product were to be used in a solar panel with an unmodified composition, this would be within the prior art as it can be used without knowing how to manufacture the commercial product.
Pete Perfect,
If you had read carefully what I have been saying all along, you could have realised that your comment is not leading us much further.
Should the claimed properties correspond to the publicly available properties of the commercial product used on its own, it is manifest that the commercial product is then novelty destroying for the claimed subject-matter. Enablement does then not play a role.
In the absence of any properties made public, or only very partial ones, it has then to be established whether the commercial product can be reproduced without undue burden, i.e. whether it can be reverse engineered. If this is the case the commercial product is still novelty destroying.
If the commercial product cannot be reversed engineered at all, then, in principle, it boils down to a non-enabling disclosure. This was the case for the resins in the present case.
In T 438/19, in Example 3 of D1 the commercial product was not used on its own, but was part of a blend with other components, and the only difference with the claimed subject-matter was the aluminium content. This content could not be determined in the absence of any knowledge of the synthesis conditions of the commercial product.
Thus, if the commercial product cannot be reverse engineered, it has to be checked whether the commercial product can be replaced by a similar product with the help of further prior art. This is the route suggested in T 438/19. The objection is then only lack of inventive step.
The above reasoning is still taking into account that, according to G 1/23, a commercial product made public before the effective date of a claim can be used against claimed subject-matter.
I would simply add not at any condition. It has to be enabled, either on its own, then the product is N destroying, or it can be replaced by a similar product, with the help of further prior art, and then the objection is lack of IS. If the commercial product is used in a blend with other components, the likelihood that the objection is lack of IS is extremely high.
A commercial product used on its own is part of the prior art, and there is no doubt about this. It however depends which property of the commercial product is used in a subsequent invention.
Daniel, I’m still not convinced.
After grant, there is “equality of arms” between patentee and opponent. EPO-PSA provides a strict and fair framework for debate about patentability. The tribunal sits back and listens carefully, then decides. Simples.
Prior to grant, it is very different. The ED is handicapped, whenever an issue of sufficiency of disclosure arises. But the scales of justice are balanced by the burden placed on applicant, to prove patentability to the Office. This is very different from other jurisdictions, where the burden is on the Examiner, to prove a lack of patentability.
I still dont see how G1/23 puts obstacles in the way of an EPO that is simply trying its best to be fair to everybody, that is to say, all of, the public, the inventor and the opponent.
Time will tell though, as you say.
Dear Max Drei,
It is your good right to still not be convinced, but it is also my good right to show that G 1/23 has left a huge gap, especially when looking at Reasons 96 in the decision.
I fully agree with you that after grant there is equality of arms between patentee and opponent. The EPO acts then as a kind of referee and issues a decision on the basis of the facts and arguments presented by the parties.
Your comment brings about an interesting point of view, not to say a problem, which can occur during examination. An ED can only launch an objection of lack of sufficiency in a very limited number of situations.
For example, lack of sufficiency can be objected to when a claimed entity is going against well-established laws of nature or only chance allows to obtain a claimed result, cf. Guidelines F-III, 3 or F-III, 4. Guidelines F-III, 4 mentions that an objection of lack of sufficiency is only possible if it is not plausible to obtain the claimed result.
An other possibility is when the invention is defined by parameters and those parameters are not measurable with well-established measurements methods or with known measuring devices, cf. Guidelines F-IV, 4.11 and 4.18, e.g. T 1305/15 or T 2096/12 as well as T 2161/17.
In T 59/08, the board made clear that if the requirement for sufficiency can be lower, the requirement for IS should be higher. The interplay between Art 83, 84 and 56 is best exemplified in the catchword of T 2001/12.
Another possibility is that after filing the measuring device is no longer available, cf. T 1293/13. In T 1293/13, the machine ‘Frazier Low Pressure Air Permeability Machine 750’ to measure a given parameter no longer existed, which led to a lack of sufficiency.
Take example 3 of D1 in T 438/19, mentioning a commercial product, e.g. Engage 8400, as being part of an application for a European patent. We know, after opposition, that the commercial product Engage 8400 is, as such, not enabled, but could be replaced by a similar product.
During examination, the ED is not at all in a position of showing the lack of sufficiency of example 3 as the ED cannot show that the commercial product, e.g. Engage 8400, is not enabled or cannot be replaced by a similar product.
The applicant might even allege an effect, possibly supplemented by experimental data under G 2/21, but the ED will not be able to check whether the effect is actually achievable, or that the commercial product is not enabled or cannot be replaced by a similar product.
When a commercial product is mentioned in the description, not even in the claims, the ED will have no choice but to grant a patent, and the EPO and third parties will have to wait for an opposition in order to properly decide about the validity of the patent. This is in my opinion highly unsatisfactory.
The scales of justice might be balanced by the burden placed on the applicant, to prove patentability to the Office, but this is barely possible in examination.
An ED cannot decide whether a commercial product is enabled or not, or if it can be replaced by a similar product, as it has no possibility whatsoever to reverse engineer the commercial product. It has the possibility to ask this to the applicant, but only if there are reasonable doubts about the non-enablement. As a commercial product is prima facie part of the prior art, the ED cannot do much, but raise doubts about the enablement.
It is therefore to be expected that the EPO might be faced in the future with lots of applications being riddled with commercial products, even mentioned in the claims, which according to G 1/23 have to be considered as belonging to prior art. An ED will not in a position to challenge sufficiency and to decide whether they are enabled or not or whether they can be replaced by a similar product. It is highly unsatisfactory to have to wait a possible opposition to clarify the situation.
Another side effect of G 1/23 is that Guidelines F-III, 7, prohibiting the use of trade marks in view of their potential lack of sufficiency, will have to be completely rewritten.
As such G 1/23 might look a big step forward, but reality is somehow different. In view of G 1/23, a commercial product is part of the prior art whether it is enabled or not or whether it can be replaced by a similar product. I see here a fundamental problem which has been completely dodged by the EBA.
I hope that I could at least convince you that G 1/23 has raised more problems than it was meant to solve.
Daniel, our discussions are somewhat assymetric because you have exhaustive knowledge of the EPO case law, and I not. Besides, I don’t know anything about chem/bio cases. But I shall continue to toss my ignorant thoughts in anyway, because I enjoy reading your replies, here to your reply to my “equality of arms” post.
How does one identify a “commercial product”? With its TM, I suppose. But a TM fails to define the product clearly. Compositions change with time and country-by-country, even while the TM remains constant. Therefore, any claim that recites a TM is inevitable an offence against at Art 84 EPC, and also very likely Art 83 EPC, is it not. How shall the applicant or patentee cure the invalidity? I have no idea. Art 123(2)/(3) trap, anybody? So, at my first glance, a plague of TM’s in patent applications strikes me as less than inevitable (at least for as long as applicants do indeed seek rights that they can indeed enforce).
Dear Max Drei,
As far as asymmetry is concerned, I do not see it as something negative, or in which you are at a disadvantage. It fosters discussions and this is positive.
A commercial product is a product which is sold under a TM as it has to be somehow identified. I agree with you that a TM fails in general to define a product clearly and that its composition might change with time. What might be disclose are some properties and fields of application.
In T 797/14 the use of a TM lead to insufficiency of disclosure. In T 522/14 the use of a TM did not lead to insufficiency of disclosure. A similar result is to be found in T 227/14.
In T 2020/13, the board held surprisingly that JEFFCAT®, used in the prior art D,1 being a registered as a trademark of Huntsman Petrochemicals Corporation, any knowledgeable reader wishing to understand the meaning of the trade names JEFFCAT® ZR-50B and JEFFCAT® ZF-10 mentioned in the same document could contact the customer service of this company, to inquire about the product sold under those trademarks and, in particular, its chemical composition.
I have reasonable doubts about this possibility. If the composition and/or method of manufacture is not disclosed in the publicly available documents relating to the product, I doubt that the holder of a TM would simply give away the composition of its products and its method of manufacture to any third party wanting to use its products.
What was more convincing in T 2020/13 is the fact that both JEFFCAT® Z-130 and JEFFCAT® ZR-50B have been added to the Chemical Abstracts database. This means that at the time of the publication of D1 via the Chemical Abstracts database, the knowledgeable reader had easy access to the information on JEFFCAT® ZR-50B, although this product was no longer sold under this name.
The Chemical Abstracts database revealed, at the time of publication of D1, the chemical formulas and chemical names behind the trade names JEFFCAT® ZR-50B and JEFFCAT® ZF-10.
In spite of what the EBA has said, a product sold under TM can be deemed not sufficiently disclosed. Sufficiency of disclosure cannot be cured without infringing Art 123(2).
It all depends whether the product sold under TM is enabled or not, or if it can be replaced by a similar product. See also my reply to Pete “Perfect”.
In spite of G 1/23, I would still avoid using a TM in a claim, unless the composition of the product is well known from another source.
I am not convinced that TM will not proliferate in descriptions of applications, as an ED will not be in a position to challenge the enablement and it puts first the burden on the opponent or the party requesting revocation of such a patent. The EBA might not have wanted to open such a door, but it has done indirectly.
Daniel, I see a contradiction in your two pronouncements found above and which I copy below:
“it’s a matter of fact that the commercial product is effectively not part of the prior art.”, whereas
“Commercially available products are certainly part of the prior art,”
Perhaps the key word here is “effectively”?
It seems to me absurd to argue that a product freely on sale on the commercial market is not, in and of itself, NOT within the state of the art, even when the producer retains as a trade secret how to make it. That product X was, on date Y, available on the open market to purchase is an unassailable fact, is it not?
By contrast, it may well be that the process of making the product is not available to the public, nor even the constitution and composition of the commercially available product. One can buy the product even while not knowing how to make it. One can try to reproduce it (or “modify” it) but it might very well be not obvious how to do that. Coming up with an “alternative” might well be a patentable invention. But hey, that’s the patent system, functioning as it should, no?
The question is of course whether product X on sale belongs “effectively” to the Art 54 or Art 56 state of the art. In general, I am not in favour of legal fictions. Some are essential (like what the “skilled person” is deemed to know. But there needs to be an over-riding reason before any particular legal fiction should be conceived.
Sorry readers. In my “absurd” sentence there are one too many occurrences of the word “not”. I’m hoping you got my meaning nevertheless.
Dear Max Drei,
There is a subtle difference which might have come out as clearly as I thought.
According to G 1/23 any commercial product available before the effective date of the claim is part of the prior art.
In T 438/19, the commercial product was not enabled, but the BA showed a route by which the commercial product could be replaced by a similar product with the help of further prior art.
In the present case, T 1044/23, the opponent did not show that the resins at stake could not be reproduced. Hence IS was given.
A contrario, it means that, if the commercial product can be reproduced, then a lack of N or IS applies. The commercial product can be reproduced as such or in a similar fashion, see T 438/19.
If the properties of the claimed product are publicly disclosed in the technical information relating to the commercial product, then the likelihood is great that N is lacking without having to know its exact composition.
As soon as the commercial product is part of a blend with other products, then the likelihood is great that the objection is lack of IS.
If the commercial product cannot be reproduced, then the commercial product cannot by part of the prior art. There is a long lasting line on case law on this.
This is why I made the distinction of the commercial product being “effectively” part of the prior art.
Daniel, sorry but for me the same apparent contradiction in your writings remains, as to whether a commercial product that defies analysis is or is not something that is part of “the state of the art”. First you say it is, then you say it isn’t.
You cite the long line of case law to the effect that only enabled sources can be novelty-destroying. Is anybody challenging that line. Certainly not me, and nobody else, as far as I know. Let us suppose that the product called Coca-Cola defies analysis. It (the product) is on the market and it has been “made available” to members of the public. They dont know what they are drinking though. The product is within the state of the art but not its composition or how to make it.
Competitors can offer altermative products but these alternatives can’t be copies or even “modifications” of the original because no member of the public knows what the original is and you can’t compare your creation with an unknown original. What you can do though is to measure a specific product performance characteristic and then see whether your alternative product measures up or falls short of the original in that specific aspect of performance.
As to whether the alternative is an improvement, indeed a better solution to a technical problem, the answer might depend on the reason for drinking the stuff. Perhaps because it performs better as cough medicine or as an oxide-stripper. Investment in alternatives to successful commercial products is a good thing and the patent system should encourage such investment.
Real cases will inevitably turn on their unique matrix of fact. Is the EBA simply recognising that reality?
Oops. I didn’t mean to say that one drinks oxide-stripper. But if I remember right, putting old coins overnight in a bath of a cola drink will strip patina from their surfaces and leave them looking newer and more shiny.
And I forgot to observe that creating an “alternative” taste of a beverage, however ingenious that creation might be, is perhaps not “technical” whereas raising cough-curing performance of the fluid might well be.
There are five (as far as I can see) potential problems, that we must not mix up. These are (+my proposed answers):
1) Is a claim citing ‘part X is made of Coca-Cola’ clear? (My answer: In general not because it may well be that the composition of Cocoa Cola changes over time/geographic location)
2) Is a claim citing ‘part X is made of Coca-Cola’ enabled? (My answer: In general not because the skilled person can not independently create Coca Cola if he can’t buy it ready-made. In other words: The contribution by the invention to the technical field is undefined (Coca-Cola is just another word for an undefined liquid composition))
3) Are the properties of Coca Cola prior art? (My answer: Yes, but only those that are analyzable and/or that are publicly known)
4) Is the composition of Coca Cola prior art? (My answer: No, but it is quite possible (depending on case-specific details) that the composition is analyzable, in which case ‘composition’ becomes a property, and then Q3 becomes relevant)
5) Is it possible to prove direct infringement? (My answer: Not based on an objective analysis of the infringing product (for lack of knowledge on the composition of Coca Cola). Yes if based on inspection of the infringer’s premises/administration. Yes for equivalent infringement)
G1/23, in my view, is all about the interplay between Q4 and Q3.
@ BJ,
I can agree in principle with your list of potential problems, with the proviso that the situation might change depending on the category of the claim. See my reply to Max Drei.
The statement “part X is made of Coca-Cola” can be objected to under Art 83 or Art 84. I am for Art 83, as clearing the objection leads invariably to added matter. Coca-cola is not enabled as such.
As far as infringement is concerned, I will refrain from any opinion, as I have never been concerned with it, only with validity.
Dear Max Drei,
I can only repeat what I have said before. There is a long line of case law saying that that only enabled sources can be novelty-destroying. There are circumstances in which the commercial product cannot be analysed and then this line of case law applies, what you clearly accept.
In the two cases you mentioned, Coca-Cola is not modified but receives a new use, either as cough medicine or as oxide or patina stripper. A new use of a known product can represent an invention.
When Coca-Cola is used as cough medicine, and its use as cough medicine has never been disclosed, this even confers N to Coca-Cola as it could well be a first medical indication!
When Coca-Cola is used as oxide or patina stripper, and if that use is not disclosed in the prior art, then the new use is novel and most probably inventive. Here it is not the product as such which is claimed, but the use of the product.
It depends thus on the claim category, the product as such can still be considered as part of the prior art, but its use is different, and the new use is the invention and not the product. If the new use is patentable, then it is indeed irrelevant whether the composition of the product is know or not.
I do howeverstill have some reservations as the formula of Coca-Cola can be modified in time and then it might for instance lose its medical indication or its effect as oxide-stripper. What then?
As soon as Coca-Cola is modified by addition of some further components, and claimed as product, we still have the problem of enablement of Coca-Cola. And this is the problem the EBA mentions in Reasons 96: “Adding some lemon juice to a glass of Coca-Cola for a less sweet taste may not be inventive simply because the recipe for Coca-Cola is secret and therefore the original taste of Coca-Cola is considered as unattainable”.
This is why I consider that even if a commercial product, is first considered as part of the prior art, when the commercial product is modified, the resulting modification might not be patentable, for the simple reason that the starting commercial product is not enabled.
I added a shortcut: what does it help to consider a commercial product to be part of the prior art if it is not enabled when modified or when the commercial product cannot be replaced by a similar product with the help of further prior art. This is the route taken in T 438/19.
What does it help, Daniel, you ask me. Perhaps I have misunderstood, yet again, but my answer is that it might help in the way that francis hagel contemplates, namely, as a contribution to the analysis under Art 56, EPC.
It amuses me, that we are now up to 37 comments because I suspect that most or all commentators have the same opinion, but the different ways of expressing it are leading to confusions and misunderstandings all round.
For example, in my strained hypothetical about Coca-Cola, I had in mind that its uses as cough medicine or paint stripper were known but that the competitor’s rival cola product would deliver a better performance than the original Coca-Cola product, for those specific uses. So, no inventive new use, just an indisputable and proven “technical effect” arising from something that amounts to a “modification” of the market-leading but nevertheless trade secret product. The claim I envisage is one directed to a cola drink, the claim specifying clearly the composition of the drink, whereby readers of the patent can reproduce both the claimed product and its improved performance, whereby the unknown composition of Coca-Cola becomes irrelevant to the patentability analysis yet the inferior performance of Coca-Cola is crucial to the analysis.
Does this thread have any more life left in it? Personally, I doubt it.
The situations considered in the preceding comments seem to relate to claims drawn to the modification of a commercially available product. If the composition and/or method of manufacture of the commercial product is not known, as is the case for the Coca-Cola recipe, it seems to me that a preliminary finding should be that the claim is not clear and/or not enabled. Shouldn’t this come before the novelty and inventive step issues ?
On the issue of inventive step, I do not understand why, generally speaking, a non-enabling piece of prior art should be excluded from consideration by a skilled person in the assessment of obviousness. A piece of prior art while non-enabling may provide teachings which provide the skilled person helpful inspiration to move to a next step toward the claimed subject matter. That coule be especially the case if the non-enabling prior art is applied in combination with a closest prior art, to show that the skilled person would have been incited to modify the CPA in a certain direction.
This could apply not just to commercial products but also to the literature. Take the case of a document which teaches the effect of a structure or method defined in general e.g. as a genus or by way of a functional definition but fails to disclose a specific credible embodiment. For that reason, it may be considered not enabling, Yet, the teaching of the document linking the general definition to a certain effect has no reason to be excluded in the assessment of inventive step, since it would be available to the skilled person and possibly effective to direct them toward the claimed subject matter.
Another consideration is that the finding that a piece of prior art is non-enabling is typically debatable and thus subject with a great deal of uncertainty, as much as the assessment of sufficiency for an application or patent.
Mr Hagel,
I cannot but agree that, if the composition and/or method of manufacture of the commercial product is not known, a preliminary finding should be that the product is not enabled. One could also argue that essential features of the commercial product are not in the claim, but if they are not in the description, then the lack of sufficiency is manifest. It is moot to discuss novelty and inventive step issues on the basis of a non-enabled piece of prior art.
Either the piece of prior art in enabling or not. If it is not enabling as a whole, then any discussion on N or IS becomes moot.
If it is partly enabled, i.e. not all its properties or its composition are unknown, and only under this condition, it can be used, as such or in combination with another piece of prior art, to assess N or IS of the subject-matter of the claim.
This applies mutatis mutandis to a piece of literature. Any teaching which can be taken from the document by a skilled person, is without doubt part of the prior art and there is indeed no reason to exclude what is known in the assessment of inventive step. Here again, the enablement is partial, and if the know properties/composition can be linked to a certain effect, preferably disclosed, then this effect is also part of the prior art.
I can agree with you that whether a piece of prior art is non-enabling is typically debatable and can be subject to a great deal of uncertainty. In examination it is, generally, not possible for an ED to raise an objection of lack of sufficiency, unless the insufficiency is prima facie apparent. The situation is different in opposition, as there the opponent has the onus of proof. If the opponent has fulfilled its onus of proof, then it is for the proprietor to show the contrary.
A non-enabled piece of prior art cannot be considered belonging to the prior art, whether it is a commercial product or not. However, the known properties or part of the composition of a commercial product are part of the prior art, but only those.
Dear Max Drei,
I understood your last but one comment in that there were new uses of the know Coca-Cola. I could not get from your statement that the uses you meant were know. This explains a possible confusion.
You had in mind that its uses as cough medicine or paint stripper were known but that the competitor’s rival cola product would deliver a better performance than the original Coca-Cola product, for those specific uses. This is a different starting point.
If the claim is directed to a cola drink, the claim specifying clearly the composition of the drink, whereby readers of the patent can reproduce both the claimed product and its improved performance, then the unknown composition of Coca-Cola becomes indeed irrelevant to the patentability analysis.
It happens however, that in G 1/23, the EBA has considered that the unknown composition of Coca-Cola is not a bar of it being part of the prior art. What is it good to state that a product of unknown composition is part of the prior art, when at the same time its composition is actually unknown, hence not enabled?
That known properties of a product, for which not all properties are known, are part of the prior art is not at stake. If not all properties are known this document cannot be regarded as CPA and is indeed useless as far as IS assessment is at stake, unless it can be replaced by a similar product, see T 438/19.
Nearly there, Daniel, I think.
Let us concentrate on the choice of CPA in the EPO-PSA approach to obviousness. I envisage cases where a successful product is the business starting point for competitors. Suppose they create an alternative product and patent it. Now, how shall the obviousness attack on that patent proceed?
In the real world, the starting point is the successful product. You say that it “cannot” serve as the starting point. I say it can. You say that the choice of such a starting point is futile because an attack based on it cannot possibly succeed. I say that it can. here’s how.
Motivated by the successful product, the skilled person underatakes a search for an enabled prior art “starting point”. But that search starts only after the skilled person is prompted to do it by the successful product already on the market. No such product, no such search. How thhen can one say that the successful product is irrelevant to the obviousness enquiry?
Dear Max Drei,
I never said that a commercial product “cannot” serve as starting point.
I did put some conditions on it:
If the commercial product can be analysed and reproduced, then it serves as starting point without any possible doubt.
If only part of the properties and composition of the commercial product are known, this known part is certainly prior art. Whether it is sufficient to render the claim obvious is a different matter. To the extent of the know properties and compositions, the commercial product is part of the prior art and starting point.
If the commercial product can be replaced with the help of a similar product, even when taking into account further prior art, cf. T 438/19, then at a pinch the commercial product can be considered as representing the CPA.
If the commercial product cannot be reproduced, then it cannot serve as CPA, as non-enabled. Case law is very clear in this respect. If the commercial product is not enabled, then it is moot to consider it belonging to the prior art and as starting point.
Thanks for that answer, Daniel. When I wrote “the” successful product, I thought it implicit that it was the non-enabled one that we were discussing. Is that not where our debate is located? As to your “moot” however, I’m still not convinced of that, for the reasons I have given. let us leave it at that, I would say.
So, let’s conclude the discussion. I bow to the weight of your deep knowledge of the EPO’s case law and accept that such a (non-enabled) product cannot serve as the CPA under EPO-PSA obviousness analysis.
I follow-up on MaxDrei’s focus on a commercial (successful) product for which the applicant is looking for an alternative. It is safe to assume in this case that the use properties of the commercial product are known. Given the applicant’s objective to find an alternative product with comparable or improved properties, the relevant information for IS assessment are the use properties. Whether the information for manufacturing the product or its composition are known in detail is irrelevant.
In the example of the patent in suit, detailed manufacturing information relates to the Ziegler-Natta catalyst used for manufacturing the PE resins. For aught I know, it is common practice for catalyst producers to protect the detailed composition or other relevant information of their products as trade secrets. At the same time, they may customise their catalyst products so as to meet a client’s specifc needs as to use properties of the resulting polymer. A commercial product is not really non-enabling just because its composition or manufacturing method is kept as trade secret, if the ingredients needed to achieve the desired result can be sourced from a supplier.
Getting back to a point discussed with Mr Thomas, the exclusion of « non-enabled » commercial products, we have agreed that this raises a great deal of uncertainty, given the usually highly debatable issue of the enabling requirement. It is so debatable esp. because of the uncertainties regarding the « art » and the skilled person and the CGK available to the skilled person. This is bound to grow with the spreading of AI tools to all fields of technology and the resulting inclusion of AI tools as part of the CGK.
A telling illustration of this trend can be found in the recent CoA UPC decision Amgen vs Sanofi, in which the « skilled person » is a team which includes a specialist of drug discovery. It is clear that the skills of such a specialist rely on the use of AI tools designed for drug discovery.
Dear Max Drei,
You have no reason to bow to my knowledge of the EPO’s case law. I have just reminded you of a long line of case law which I do not think having been superseded by G 1/23, may be completed.
I do not think that we are fae away from one another. It depends on the category of the claim: composition or use. What is know from a commercial product undoubtedly belongs to the prior art and can be used against a claim irrespective of its category.
That’s were the difficulties start.
Dear Mr Hagel,
As I said to Max Drei, we are not far away form one another.
Anything known from a commercial product, e.g. its properties belongs to the prior art. This is why claim 1 as granted was lacking N as the known properties matched the claimed ones.
The problem came up with AR4 as then the properties claimed did not match those of the claim, and hence IS came into play.
If it can be shown that the commercial product can be replaced by a similar product, then IS is lacking, not because of the commercial product as such, but due to the fact that a similar can be developed, cf. T 438/19.
In the present case, the gap was too big and the opponent could not show that either the commercial product or a similar could be produced, hence IS was given.
When a product is developed with the help of AI, the product might be known, but not the way to obtain it. This is a first problem. A second problem is whether the data used by the AI was itself public before the effective date of a patent.
That the skilled person can be a group of persons is an acquired fact at the EPO. I used to work for a long time in pacemakers. A pacemaker needs a doctor (if the pacemaker does not correspond to the physiological needs of a patient it is useless), an chemist (there are electrodes to be insulated and interfaces with tissues and blood), an electronic engineer (to program the device) and a mechanical engineer (for the casing).
We did not have to wait for the UPC to come to this conclusion.
Mr Thomas,
I am pleased that you find that we globally agree.
As to the recent UPC decision I cited, please do not misunderstand my position. It is neither to automatically approve any decision of the UPC court of appeal, nor to say that the EPO BOAs and EBA must strictly align with the UPC. Convergence is desirable, but as the contexts are different, the requirement for a strict alignment is not warranted.
The CoA UPC decision Amgen vs Sanofi of 27 November is a case in point. It seems to me the definition of the skilled person as including a specialist of drug discovery is to the very least disputable. Is really » drug discovery » an area of technology as such ? The implications as to the assessment of inventive step are a concern. What exactly will be the skills of a drug discovery specialist, compared with those of the conventional skilled person as defined in the case law ? It will be interesting to see how colleagues of the pharma sector react to this decision.
Mr Hagel,
If everybody would always agree on everything, life would be very boring.
I think that the EPO and the UPC will attempt to come to a kind if modus vivendi when it comes to the definition of the skilled person or validity, but I doubt that they will ever be aligned. The Amgen-Sanofi decision you have mentioned is a good example. Some of the terms used by the UPC CoA appear to sound familiar to people used to the PSA, but have a different meaning in the two instances.
Civil courts will always have to define the skilled person and “interpret” the claims, as they do not have the technical knowledge to decide on the precise meaning of each claimed feature. This is a matter of fact and has to be accepted.
The EPO technical boards of appeal, do not need to delve so deep in interpretation of claims and definition of the skilled person, so that do only rarely need to define the skilled person, they can decide more directly on what is common knowledge of skilled person.
This might not have come over so well in my comment, but I hope the situation is now clearer.
As the addressee, the “skilled person”, does not exist in real life, we can’t ask them what the claim means, to somebody like them. Somebody else has to tell us.
A 3-person EPO tribunal can do it quite well, relying on their own deeptechnical knowledge, surely better than a single non-technical judge. But what about a single judge who has spent the best part of a week noting how each of the opposing sides’ technical experts reply to rechnical questions about the few points of contention, put to them (through the advocate) by the other side’s technical expert, in the witness box, under intense cross-examination. Two world-famous professors of chemistry, cross-examining each other over some tiny but decisive point of detail, over a couple of full days in court, is a not-to-be-missed experience.
The outcome is a statement by the judge what the skilled person knew and what they did not, a statement which is so authoritative that neither side can find fault with it. If the judge gets it wrong, it’s because one party or the other (or both) find out with hindsight and regret that they unfortunately chose the wrong technical expert. OK, the English system is inherently expensive, but it can lay down a roadmap that saves the parties litigating in all other jurisdictions. In some cases, the expense of the trial in England is proportionate.
And it’s relatively quick too. Less than 2 years to an exhaustively reasoned written decision which, in most cases, is pointless to appeal.
How will the UPC decide the decisive question, what the skilled person thought and knew, back at the date of the claim? Who knows how they will do it. What are their options? Or will the international advocacy firms set the course? As Chou-en-lai said (or was it Mau Tze Tung) of the consequences of the French revolution “It’s surely too early to tell”.
Dear Max Drei,
I can agree with you, that the skilled person does not exists. At least in my 50+ years in the patent business, I never shook hands with one.
What you actually describe corresponds to the well-known cross-examination system carried out in English courts.
I agree with you that the result of an English cross-examination system is probably quite waterproof. The question is however at what price, and yet it still needs nearly two years.
This is why I think that English courts might still play a role in the future, but since no jurisdiction will freely accept a decision from a foreign court, and certainly not from a court outside the EU, this way of doing is deemed to perish, like national court actions.
When ratifying the UPC, parliaments have not been made aware of the real possibility that their own judiciary in patents could become obsolete due to the UPC. It is not for no reason at all, that Germany and France have accepted a double patenting, national and European patent with unitary effect.
It was clearly not the way the UPC was conceived from the outset. It should be quick, but not cheap, as even the procedural fees will increase soon.
In view of the speed at which the UPC wants to come to a decision, errors will inevitably occur. As the UPC is a court made for big players with deep pockets, supported in general by internationally active litigation firms, the odd wrong decision will not matter for the big players.
This is one of the reasons, I cannot be in favour of the UPC, as it is not a court made for SMEs, in spite of all what has been heralded by the promoters of the UPC. The contrary is true.
This is also a reason why oppositions at the EPO could be a good way to check the validity of a patent before starting an action before the UPC. As a lot of the first instance decisions are set aside by the boards of appeal, it is necessary to go to appeal, unless the case is clear cut.
The ideas of the opposition is not my idea. It has been brought forward by a British representative during a conference in which Sir Robin Jacob wanted to hear that the EPO boards of appeal will follow the decisions of the UPC.
For Sir Robin Jacob, the UPC should become the leading court in patent matters in Europe. This was before the London Section of the CD and the London Local Division were declared dead and fiddling began at the UPC.
For your information, I will shortly issue a blog dealing with IS as seen by the CoA UPC. in view of the way claims can be interpreted, divergent decisions will occur, at least until the effect of G 1/24 has not diffused into all applications filed at the EPO.
In a nutshell, forget the PSA, when it comes to assessing IS.
Daniel, you mention Robin Jacob. He it was, who liked to crack a joke at the start of his address on patent law, for example by telling his audience to note well what the Americans did (pause) and then to be sure not to make the same mistake. Ha Ha!
Below is a link to a decision by the US Court of Appeal, 2 December 2025, on a case where a jury decided whether there was any added matter in the patent. The jury said no but the court of appeal ruled that no sane jury could so decide. The mind boggles.
The US patent involved (10,808,039) has at least 8 (eight) family members at the EPO. I haven’t looked at any of them.
Not a good idea, I would say, in Europe, to have a jury decide what the person skilled in the art knows or thinks.
https://www.cafc.uscourts.gov/opinions-orders/23-2424.OPINION.12-2-2025_2612088.pdf
Dear Max Drei,
I can agree with Sir Robin Jacob, when he says that in Europe we should not repeat the errors made on the other side of the Atlantic.
However, I do have to disagree with him, and Lord Neuberger, when they both claim that the examiner was manifestly mistaken when he raised an objection under Art 123(2) in the famous “pemetrexed” case. The examiner had no choice in view of the original disclosure. To use an US expression the applicant/proprietor was not in possession of the broad invention “pemetrexed as such and all possible salts” when the application was filed. He only disclosed one salt.
Looking at the case in the US, that a jury can decide a 41 millions $ of damages is flabbergasting. That the Court of Appeal had to step in is also quite remarkable. A jury deciding on patent matters is certainly not something we want to see on this side of the Atlantic.
Filing in July 2019 a continuation of an application originally filed in 2004 and claiming that they possessed the later invention already in 2004 is already quite daring, but on top assigning an alleged infringer in 2020 is even more daring.
I have always been very weary when I saw “continuation” or “continuation kin part”. I barely remember a case in which it was correct what the US applicant claimed about its continuation or continuation in part. .
Daniel, you write “forget PSA”. In that context, I have just read para 8 of the appeal instance of the UPC in Meril vs Edwards:
https://www.eplaw.org/blog/detail/upc-edwards-v-meril-court-of-appeal/
In that paragraph, you will see the word “compare”. Because I cannot find anywhere in that paragraph the complementary word “with”, I do not know with what other document I must “compare” the claim in view.
In other words, what the UPC calls the “objective” technical problem looks to me like the EPO’s “subjective” technical problem, ie not “objective” at all.
Can you, or perhaps francis hagel, help me to understand UPC thinking on Art 56 EPC? For example, perhaps you can point me to a different thread on your ever so informative blog. Thank you.
Dear Max Drei,
When I said forget the PSA, I meant it in the present case. If the prior art is not enabled, then you can forget any application of the PSA.
A far as the Meril case at the UPC is concerned, I published today on my blog an analysis of two decisions of the UPC CoA and tried to show were are differences and were are similarities when assessing IS. One thing is clear, the UPC CoA has not adopted the PSA. .
My conclusion is exactly yours the “objective problem” at the UPC is the “subjective problem” at the EPO when applying the PSA. It can well happen that the OTP at the EPO corresponds to the subjective problem, but this is in general only the case if no better PA is found during the search.
Using the same term at the UPC and at the EPO, i.e. the “objective” problem, will most probably lead to some confusion for the not so attentive reader. I do not think it will help harmonisation.