The decision of Mr Justice Coulson in the Technology & Construction Court in Triumph Controls, follows a line of cases, including the often referenced judgment of Morgan J in Digicel, where parties have been reprimanded and sanctioned for taking unilateral decisions in disclosure exercises, failing to follow the rules and cooperate with opponents.
That the Claimants' failures in Triumph Controls concerned the use of some relatively new technology and techniques, was really incidental to the case's central issue of failure to comply with the Civil Procedure Rules.
The court considered the Defendants' application for further disclosure. It was supported by evidence highlighting deficiencies in the Claimants' approach to disclosure.
The court noted that disclosure had "been something of a running sore in this case". Against this background, the court had to consider whether to grant two orders sought by the Defendants for further disclosure:
- a first order that a list of 860,000 folders and file paths located on a shared drive should be provided to the Defendants by the Claimants, so that the Defendants could verify whether there were any folders or file paths that should have been searched for disclosure (Issue 1); and
- a second order requiring the Claimants to manually review 220,000 documents which had been determined as potentially disclosable but only reviewed by a limited sampling exercise (Issue 2).
In an attempt to limit the disclosure process to manageable volumes, the Claimants had asked particular custodians to highlight which of the 860,000 folders or file paths may be relevant for disclosure, and their solicitors conducted a cross check of the full list. This reduced the sample reviewed by two thirds, from 3 terabytes to a single terabyte of data.
The selection process was conducted without first consulting the Defendants. However, the court considered, that while it would have been better had they done so, the approach taken by the Claimants was acceptable because they had clearly outlined the process of their selection in their original list of documents served on the Defendants. Further, despite having the opportunity, the Defendants were not able to identify any potentially relevant file paths or folders which were obviously missing.
The Judge considered that the Claimants had always been clear about their proposed approach to the shared drive. Any complaint raised by the Defendants about it now was very late and the evidence showed that the Claimants' approach had been reasonable and proportionate. The Defendant's application on Issue 1 therefore failed.
Of the documents initially searched using key words, 450,000 were identified by the Claimants as potentially disclosable. The Claimants manually reviewed the first 230,000 of these documents, assisted by what was described as "computer assisted review" , a process referred to as Technology Assisted Review (TAR) in this article. At no point did the Claimants provide a clear explanation of how this review was undertaken.
In respect of the remaining 220,000 documents, the Claimants were said to have reviewed a 1% sample (i.e 2,200 documents) using TAR, which led them to the prediction that only 0.38% of those documents would be relevant to the case. The Claimants decided (unilaterally) that it would therefore be disproportionate to continue reviewing the remaining documents.
The Judge found that this approach was inconsistent with the approach first outlined in the Claimants' Electronic Documents Questionnaire (which had only specified a "manual review").
After the deadline for disclosure, the Claimants disclosed some 3,000 further documents (largely because the Claimants' experts and witnesses sought to rely on them). The Judge noted that this in itself demonstrated deficiencies in the selection process, and cast doubt on the Claimants' prior prediction that only 0.38% of the 220,000 previously un-reviewed documents were disclosable.
The court also criticised the Claimants' unilateral decision to use TAR, without discussion with the Defendants, and without any explanation as to how the TAR was conducted, or set up. The court was left speculating as to the reasons that the review had failed to identify relevant documents and commented that the process the Claimants followed was not "transparent and cannot be described as independently verifiable". Touching on some important issues regarding the transparent use of TAR the court noted:
"…there is no information as to precisely how that sampling exercise was conducted. There are, for example, no stated tolerances and no explanation of how many rounds of sampling were undertaken. That again is unsatisfactory. It is not in accordance with the TCC e-disclosure protocol. I am bound to say that, given that both the [TAR] and the sampling were unilateral, I am slightly surprised that there is not now better evidence as to what actually happened. There is always a risk, as Digicel makes plain, that a unilateral decision will be carefully scrutinised by the court at a later date, and a different course may be ordered. That will be more likely if the evidence as to what was done remains vague.
It was held that although an order for a further review was justified, in the interests of proportionality, and time constraints, the full review of the 220,000 documents should not be ordered. Instead, the parties were ordered to agree an approach by which a sample of 25% of the 220,000 should be manually reviewed within three weeks.
The issues in this application highlight the fact that some of the proposed new disclosure rules (to be piloted shortly in the Business and Property Courts) are necessary, in increasing the pressure on parties to cooperate to flush out disclosure issues early and agree a proportionate process for the search and review of disclosable documents. The proposed rules will force parties to discuss TAR in cases like this one, in which there are more than 50,000 reviewable documents.