T1465/23 – No Narrowing by Description—EPO Board Terminates Inventive Step Analysis for Arbitrary Modifications Citing G1/24 and G1/19

“The potential patentability of a specific narrow embodiment…cannot render a claim allowable which, due to its breadth, encompasses a multitude of other, non-inventive embodiments” – r. 3.5.

Background

EP3113515 was granted with the following independent claim:

1. A hearing device comprising:

a processing unit configured to compensate for a hearing loss of a user of the hearing device;

a memory unit; and

an interface

characterized in that the processing unit is configured to:

• receive a connection request for a session via the interface;
• obtain a session identifier
• transmit, via the interface, a connection response comprising a hearing device identifier and the session identifier;
• receive, via the interface, an authentication message comprising an authentication key identifier and client device data
• select a hearing device key from a plurality of hearing device keys in the memory unit based on the authentication key identifier;
• verify the client device data based on the selected hearing device key; and
• terminate the session if the verification fails.

During opposition proceedings, the Opposition Division held that Claim 1 differed from the closest prior art D1 (EP2760225) in all of the method steps performed by the processing unit (steps (a) to (g)). The Proprietor argued that in particular, steps (d) to (f) allows the hearing device to derive different levels of authentication in the hearing device for client devices such that different client devices may be supported in a scalable and secure manner, without needing to exchange secrets in plain text during the session setup. As such, it was argued that the technical contribution was the provision of  a device that is less vulnerable to attacks.

The Proprietor argued, and the Division agreed, that D1 taught away from features (d) to (f) as the hearing device in D1 sends a challenge to a specific “presence device” and thus knows from the outset the public key of the presence device. There is thus no motivation to include an authentication key identifier as part of the message such that Claim 1 was deemed inventive.

The Appeal

During the appeal procedure, the Board followed the Proprietor and Opposition Division to the extent that features (a) to (g) were novel. However, the Board disagreed that the technical effect relied upon by the Proprietor (i.e. that different levels of authentication is provided) is present across the entire scope of the claim.

In fact, although features (a) to (g) were linguistically connected by the antecedent basis of the features, the Board was inclined to follow the Appellant Opponent’s position that the steps were functionally independent from each other, amounting to nothing more that “islands of cryptography”. As such, the Proprietor’s alleged technical effect (increased  security) resulting from the aggregation of the steps could not be acknowledged across the entire scope of the claim.

For example, the Board held that feature (d) does not require that the transmitted “hearing device identifier” is required to be the device that is actually performing the method. Rather, the hearing device identifier could belong to another hearing device in a binaural system. This would also be technically feasible as it means both devices don’t need to be independently authenticated.

Similarly, the Board held that, contrary to the Proprietor’s assertions, the term “based on” does not require a direct or causal link between the “authentication-key identifier” and the selected “hearing device key”. Instead, the Board found that the selection could be mediated by a series of indirect look-ups such that there is no guaranteed or technically meaningful constraint between the identifier received and the key selected.

As a result, not only was Claim 1 found not to explicitly require different levels of authentication, but such an effect could not be considered to necessarily result as, when interpreted in accordance with G 1/24, the scope of Claim 1 covered embodiments that did not solve the problem of protection of modification attacks through hierarchical authentication levels.

The impact of G 1/24 across recent Boards of Appeal decision, including how it has been applied by the Board in this case, is discussed in an article found here.

Using the description to interpret the claims – G 1/24

The Proprietor argued that G1/24 should be used to interpret the claims in such a way that the specific embodiments disclosed limited the scope of Claim 1 to embodiments that addressed the aforementioned objective technical problem of hierarchical authentication levels.

The Board rejected this argument, instead stating that they had duly “consulted” and “referred to” the description when construing the claims, in accordance with G1/24, but that the description cannot override a stable and well-understood meaning. The Board therefore held that G1/24 does not afford the Proprietor an opportunity to artificially circumvent patentability objections based on prior art by merely interpreting the claims more narrowly than is reasonable based on the ordinary meaning of the terms. The Board also noted that the a similar interpretation of G 1/24 was found by the Board in T 1999/23 (also discussed in the article linked above).

Applying G 1/19

Instead of the Proprietor’s objective technical problem, the Board found that, following the principles of G 1/19, no technical effect could be formulated because the distinguishing features do not credibly achieve any technical effect across the whole scope of the claim. As a result, the Claim provides nothing more than an “arbitrary modification of the design of a known subject-matter which, being arbitrary, cannot involve an inventive step”.

The Proprietor argued that the principles of G 1/19 were limited to computer implemented inventions. The Board rejected this argument as decisions under the Enlarged Board serve to ensure uniform application of the law concerning matters of fundamental importance. The notion that the decision would thus only apply to a subsect of technical fields could not be followed. The principle arising from G 1/19 that the technical effect achieved must arise across the full scope of the claim based on the common general knowledge is therefore applicable to all inventions.

Thus, as in the present case, once it has been appreciated that there is no technical effect across the full scope of the claim, there are two options available:

1. amend the claim to exclude embodiments not achieving the desired technical effect; or
2. reformulating the objective technical problem to be less ambitious.

However, if there is no technical effect, it is unnecessary to artificially formulate an (unsolved) objective technical problem, such as finding an “alternative way to achieve a (non-existent) technical effect.”

Commentary

When distilled, there are two main points that arise from this decision:

First, it is often assumed that the greater the number of distinguishing features, the more likely it is that a finding of inventiveness will result. However, by applying the principles of G 1/19, the Board have resolutely reaffirmed the principle that a mere aggregation of arbitrary features cannot result in a finding of inventiveness.

As such, the EPO have affirmed a seemingly lower threshold for obviousness that eliminates the possibility for relying on the would/could approach if there is no technical effect attributed to the distinguishing feature. In other words, when there is no attributable technical effect, it is seemingly irrelevant whether the skilled person would make the modification leading to the distinguishing features, or whether there is a prior art teaching towards the modification specifically, instead, the only thing that matters is whether a technical effect is achieved. If not, the modification is arbitrary and cannot contribute to a finding of inventiveness.

As such, the importance of linking each fall back position to a causally linked effect during drafting is undeniably critical to provide flexibility to amend to overcome prior art disclosures to focus in on a desired and achieved technical effect.

Second, the Board rejected the proposal that G 1/24 provides a Proprietor with a means to narrow the interpretation of a broad claim with a view to artificially generating a technical effect across the whole scope of the claim.

This is potentially the most impactful takeaway for practitioners as it reinforces the primacy of the Claims and that the “broadest reasonable interpretation” and “ordinary meaning of terms” should be predominant.

When considered together, this decision highlights the tension between a Proprietor’s right to seek a broad scope of protection around their invention and the requirements of the EPC that inventions must be inventive, and not be merely arbitrary modifications or selections of known principles.

This tension may be most notable in the Board’s conclusion on the term “based on”. Many practitioners would see it as reasonable to argue (, similar to the Proprietor,) that the term would provide a causal link between the subject and the object straddling the term.  The Board’s view that such an interpretation is too narrow highlights the importance of ensuring that there are appropriate fall back positions provided in the specification that explicitly and unambiguously provide a desired link or effect under the broadest reasonable interpretation.

This article was prepared by Senior Patent Attorney Alexandra Tyson.