Diverging case law on claim interpretation less than a year after G 1/24? (T 439/22)
In June 2025, the Enlarged Board of Appeal (EBA) resolved one of the longest-standing debates in European patent law: must the description always be consulted when interpreting claims for patentability? The answer — a clear “yes” — was intended to bring uniformity. Less than nine months later, however, a new divergence has emerged. While the Boards agree on the obligation to consult, they disagree on what should follow from that consultation.
1. The finding of G 1/24
The EBA’s Decision of June 18, 2025, held that “the description and drawings shall always be consulted to interpret the claims … and not only if the person skilled in the art finds a claim to be unclear or ambiguous when read in isolation”. The decision left open what “consulting” should entail in practice, referring to “the existing body of case law from which the applicable principles of claim interpretation can be extracted” (Reasons 10). This task was left to the Boards of Appeal to define, and the EBA’s reluctance has already proven to be the source of renewed divergence.
2. Two incompatible approaches
By now, more than 80 decisions have been rendered that cite G 1/24. Many of them deal with arguments of the proprietor in favor of a narrow claim interpretation based on embodiments outlined in the description which, clearly, were bound to fail. However, two incompatible approaches seem to have emerged with regard to the consideration of definitions in the description and the reading of claims in view of the patent as a whole.
2.1 The holistic approach
T 439/22 (December 11, 2025), the referring case itself, adopted the broadest reading of G 1/24. The claim contained the term “gathered sheet”, understood in the art as folded/convoluted. But the description defined it more broadly — including sheets “compressed or constricted substantially transversely to the cylindrical axis of the rod”. The Board held that claim interpretation is “the result of both reading the claims and consulting the description and drawings as a unitary process” (Reasons 2.4, headnote 2) and that the skilled person will try to take a definition “at face value … taking into account both the broadening and limiting aspects” as “long as the definition is technically reasonable and complies with the overall teaching of the claims” (Reasons 3.4, headnote 3). The broader definition destroyed novelty.
T 1849/23 (September 2, 2025), from the same Board as T 439/22, concerned a claim requiring an angular rate sensor positioned and configured to measure the rate of angular trailer deflection about a hitch pivot point. Strictly read, this implied direct measurement. The description, however, disclosed only a gyroscope measuring absolute yaw rate, approximating trailer deflection under certain conditions. Applying G 1/24, the Board broadened the interpretation: claim interpretation requires taking into consideration the wording of the claim and the content of the description (Reasons 1.3). The claim was construed as requiring measurement under certain conditions, again rendering the prior art novelty-destroying.
2.2 The claims-primary approach
T 2027/23 (June 30, 2025), decided only twelve days after G 1/24, articulated a markedly different methodology. The Board identified three “takeaways” from G 1/24: the claims are the “basis”, and the description and drawings should always be consulted or referred to; discrepancies with national courts are undesirable; and the proprietor must remedy claim-description misalignment (Reasons 3.5.2).
Particularly, concerning the first takeaway, the Board noted that the extent to which the description must be “consulted” cannot be inferred from G 1/24, and offered four interpretations: “consulted” for the purpose 1) of deriving explanatory aids to claim interpretation, 2) of confirming the ordinary meaning of a claim feature, 3) of determining the function or the purpose of a claim feature, or 4) simply defining the technically skilled reader of a claim.
The Board then proceeded to “consult” the description in detail. At Reasons 3.5.10, it stated: “As a first step, the board has ‘consulted’ and ‘referred to’ the present patent description and drawings to define the technically skilled reader from whose perspective or view point a claim is to be interpreted”. This clearly favors approach 4) outlined above. Nevertheless, the Board then examined each disputed claim term against the description (Reasons 3.5.11–3.5.14), finding in each case that the description did not justify a narrower reading.
This Board’s approach, giving the description at most a minor weight, can also be seen in the decision T 837/24 (December 8, 2025). Therein, it gave a claim term a meaning based on a theoretically possible linguistic interpretation of the claim wording. This interpretation was contrary to the sole understanding when taking into account the description and reading the patent as a whole. The patent was revoked due to added matter which the Board identified based on interpreting the claim at odds with the description, creating an inescapable trap for the proprietor.
T 1999/23 (July 18, 2025) outright rejected limiting the claim wording in view of a restrictive definition in the description (Reasons 5.6–5.7):
„eine einschränkende Begriffsdefinition in der Beschreibung [darf] nicht dazu verwendet werden […], den ansonsten für die Fachperson eindeutig breiteren Anspruchsgegenstand einzuschränken“.
In English:
“A restrictive definition of a term in the description [must] not be used [...] to limit the scope of the claim, which is otherwise clearly broader for the skilled person”.
This principle remains valid after G 1/24, as interpretation starts with and is based on the claims, according to the Board. The Board added a legal-certainty rationale: the proprietor had created the inconsistency without identifiable justification, and resolving it against the proprietor reflects the primacy of claims confirmed in G 1/24 (Reasons 5.9).
This decision at first sight appears to be in contradiction with T 439/22, which held that definitions need to be considered at face value both when broadening and limiting the claims. However, this contradiction might be resolved, as it seems doubtful whether the definition relevant in T 1999/23 would have passed the test outlined in T 439/22 that it is “technically reasonable and complies with the overall teaching of the claims”.
3. Practical advice
G 1/24 has not created a safe haven. There is still considerable uncertainty concerning claim interpretation which applicants should take head-on. During drafting and prosecution, claim language should be used that is as unambiguous as possible. Using the description as a dictionary to define claim terms in a peculiar way remains a risk. In case of doubt, broad terms should be used in the claim that are backed up by clear narrowing definitions in the description that are available as fallback positions for claim amendments, if needed during opposition. Particularly before amending claims, all literal readings an amendment may entail should be considered to avoid a reading that could lead into an inescapable trap.
4. Outlook
G 1/24 settled whether the description must be consulted. By leaving open how and to what end, it has sown the seeds of a divergence that may — sooner or later — prompt a further referral. A larger number of decisions currently seems to favor the claims-primary approach, with three Boards aligning through express citation chains (T 1402/24 → T 2027/23; T 78/24 → T 1402/24 → T 2027/23). But Board 3.2.01’s holistic approach is arguably closer to G 1/24’s spirit, and it remains to be seen whether further Boards will expressly favor that approach.
Whether a further referral — this time on the consequences of consulting the description — will prove necessary is an open question. For now, claim interpretation at the EPO remains, to a material degree, a function of which Board decides.
