T 0832/22: Determining public availability of internet publications - standards of proof

Patents EPO Case Law

The Board of Appeal [the Board] focused on the standard of proof required to establish public availability. The Board assessed, based on a digital forensic report submitted by the Opponent-Appellant, whether a technical manual had been made available online before the patent's priority date and whether the standard of "beyond reasonable doubt" was met. The Board concluded that the public availability was proven beyond reasonable doubt.

Background

In T 0832/22, the patent related to a regulation method for an inverter compressor in a refrigeration system. During opposition, one of the key discussion points was whether a technical manual (D9) titled "Heos - High efficiency showcase controller" had been made available to the public before the priority date of the patent. The Opposition Division rejected the opposition.

The Party’s arguments

The Opponent-Appellant submitted a digital forensic report (D39) and several affidavits to argue that D9 had been published on the internet before the priority date.

The Patentee [P] contested the prior art status of D9, arguing that it was insufficiently proven that D9 had been published on the internet before the priority date. P claimed that the evidence presented was contradictory regarding the publication date. Moreover, since D39 came only from the sphere of the Opponent-Appellant, it was not possible to reduce the standard of proof to the “balance of probabilities" but instead the higher standard "beyond reasonable doubt" had to be applied. P argued that the higher standard was not fulfilled due to the contradictions around the publication date of D9 in the various pieces of evidence.

The Board’s considerations on the standard of proof

Regarding the standard of proof, the Board referred to G 1/12 and G 2/21 to note that proceedings before the EPO were to be conducted in accordance with the principle of free evaluation of evidence and that this principle required the assessment of evidence according to a judicial body's own conviction, without observing any formal rules.

The Board explained that the EPO standard of proof is generally the “balance of probabilities”, and only by exception the standard of "beyond reasonable doubt’’ is applied. The latter occurs mainly in opposition proceedings when for example only the opponent has access to evidence concerning an alleged public prior use. The Board referred to the Case Law of the Boards of Appeal, 10th edition 2022, III.G.4.3.

The Board found it difficult to quantify the difference in the required degree of conviction between the "balance of probabilities" standard and the "beyond reasonable doubt" standard and found that an attempt to describe this difference by numerical thresholds, e.g. as certain percentages of likelihood, could be misleading. Regarding the "beyond reasonable doubt" standard, the Board found it more expedient to focus on the term "reasonable". The "beyond reasonable doubt" standard did not require absolute certainty but could be met if the remaining doubt can be understood as overall insubstantial in view of the entirety of the available evidence.

The Board noted that if the deciding body was convinced beyond reasonable doubt that an alleged fact occurred, there was no need to decide which standard of proof is applicable (referring to T 2466/13, Reasons, 2.1.1), because if the higher one of two disputed standards of proof was met, it could be left open which of these standards must be applied. The Board found that this was the situation in the present case.

The Board’s view on the evidence for the present case

The Board found that the digital forensic report D39, filed by the Opponent-Appellant, established that D9 was publicly available on the internet on 19 November 2014, which was well before the priority date of 29 March 2017. D39 documented the upload dates of documents from the opponent's computer system to a public web server. The Board found no reason to question the report's conclusions regarding the 19 November 2014 upload date, noting that P had challenged neither the credibility of the report nor its methodology.

The fact that D39 is commissioned by the Opponent-Appellant and is based on digital evidence from the Opponent-Appellant has, in the absence of any inconsistencies or other counterevidence, no negative impact on its credibility."

The Board -

D39 concluded that a document named "+0300078EN" in version 1.0 was uploaded to the internet on 19 November 2014 at 13:00. According to the Opponent-Appellant, this document corresponded to D9; in response to which the Board noted this was not challenged by the P. The data in D39 concerning this document (title, release number, and date) matched information found in D9. Additionally, D39 indicated that version 1.0 was last updated by a person on 19 November 2014, which was consistent with the person’s affidavit (D38) stating she had personally uploaded the document to the website on that date.

The Board addressed alleged inconsistencies in other evidence but found that such inconsistencies were not sufficient to cast doubt on the validity of the conclusions drawn based on D39. The Board further found that inaccuracies in testimonies could be explained by the time elapsed since the events, such as confusion between a document's finalisation date and its actual upload date. The Board argued that even if, for the sake of argument, it had to be considered which of the contradicting publication dates in documents D9' or D39 was the correct one, both of them were public well before the priority date of the patent.

Decision of the Board

The Board concluded that D9 was part of the state of the art, as its public availability before the priority date was established beyond reasonable doubt. In the end, the decision under appeal was set aside and the patent was revoked due to lack of inventive step.

Summary written by the NLO EPO Case Law Team