Standard of Proof in Determining Accessibility of Online Prior Art

Standard of Proof in Determining Accessibility of Online Prior Art

Alex Gloor is a JD Candidate at Osgoode Hall Law School

A recent decision regarding a UK patent application by Ranger Services Ltd raised an interesting issue regarding the accessibility of online documents for prior art searches. As was first discussed on the IPKat blog, the specific question for discussion is the appropriate standard of proof to be used in determining the date on which internet-hosted prior art is deemed available to the public.

This case is closely related to my recent post discussing the accessibility standard of prior art searches in light of the In Re Lister decision. The future of Lister seems destined to come down to a determination of the date that a manuscript was posted on two online databases. As such, it will be relevant to the interested parties to determine whether the standard of proof for determining the date of posting is a basic balance of probabilities or the elevated standard of beyond a reasonable doubt.

As acknowledged by the EPO, online information can pose various problems for prior art searches. First of all, websites are often updated without any record of their previous content. Second, websites do not always make available the date that content was published. Third, many sites may be password protected or may require payment for access. Lastly, information on web pages can easily be manipulated to deceive viewers. Despite these difficulties, virtually all forms of internet material including password protected sites, blogs or even email archives are valid sources of prior art.

While it may seem obvious that the civil standard would apply, the applicant contended that the traditional EPO practice is to inquire on the reasonable doubt standard. In actual fact, the stated EPO position is somewhat of a hybrid, requiring more than "mere probability" but not reaching the standard of reasonable doubt. Regardless, the Hearing Officer determined that the standard to be applied in the UK was the customary balance of probabilities test. It was noted that the binding authority is UK precedent, not the EPO position.

To the extent of my research, this issue has not been commented on in the Canadian context. However, the 2003 Patent Board decision of Industrial Design Application No. 1998-2666, Re 25 C.P.R. (4th) 373 does provide guidance on the use of the internet as a tool for prior art searches. In particular, it is stated that examiners "must be able to determine the date when the design was published on the Internet" in order to classify the information as prior art. However, it is acknowledged that despite the "transient nature" of the internet, examiners will use the date stated on the website as that of the most recent modification.

The practice mandating that the date "must" be ascertained but allowing examiners to simply take the date posted at face value seems open to attack. The situation is foreseeable where an interested party changes the date on their website to falsely represent their publication so as to preclude a competitor's product from patentability. Likely the most fair interpretation of this passage would be that prima facie an examiner must show beyond a reasonable doubt at least the posted date of the prior art, with this evidence open to a rebuttal on a balance of probabilities from the applicant that proves otherwise.

A final issue deals with material that at one point existed online as prior art but was subsequently modified. Generally, internal archives are not maintained that contain a site's history. However, some third party websites such as, claim to do so. This site was in its infancy in 2003, and while it was considered in Industrial Design, no concrete ruling was made on its applicability. was considered more extensively in a 2007 European patent application and rejected as being unreliable. Factors considered were numerous, notably that was funded by the for profit, that archived material may lose its functionality, that the format of the archived site may be different to the original, that "images" of a site represented a series of snapshots rather than a continuous history and that there are proprietary issues of the information received. Given this abundance of issues, material archived by a third party without any proof of veracity is unlikely to be accepted.

The use of online material as prior art raises many interesting issues. The UK decision in Ranger Services likely was correct in applying the civil standard of proof, but the status of the law is still uncertain in Canada. However, this issue may not remain stagnant for long as it is certainly possible that cases will soon come forward specifically addressing this issue.