The review of key documents is an integral step in the discovery and preparatory stages of litigation. This task has become increasingly costly and more time consuming for clients given the increased volume of data generated by electronic communication and online business practices. However, technology assisted review tools in the area of electronic discovery such as predictive coding, have been championed in other jurisdictions for significantly accelerating the process and reducing costs.

There are initial signs that the technology is gaining traction in Ireland with Counsel for IBRC in its action against Quinn Insurance Limited and others, requesting the Court’s permission to use “technology assisted discovery” to analyse the relevance of 1.75 million documents. The Court has allotted a two day hearing, to commence on 16 October 2014, to determine whether technology assisted review should be permitted and if so, whether all parties should be allowed to have an input into its use. 

The predictive coding system operates by assessing a small set of documents that have already been analysed by members of the core legal team.  The system will then determine how likely it is a document is considered relevant based on the previous assessments made in respect of the reviewed set.  This process results in the most relevant documents being available for review at the earliest opportunity, leading to an early consideration of the key issues and an expedited review process.  In this regard, it is important that the initial review must be carried out by lawyers with an intimate knowledge of the case. While it has been argued that the technology potentially can be entrusted to perform the entire review, in practice a manual human review ought not be entirely excluded from the process.

Up until the Quinn litigation, predictive coding had never been considered by the Irish courts.  However, experiences in the US show a marked acceptance of the technology.   In the breakthrough case of Da Silva Moore, it was judicially endorsed as being superior to other discovery alternatives, such as keyword searches (which have long been accepted by the Irish courts).  Anecdotal evidence suggests significant cost savings for cases involving large volumes of discovery documents, and research shows that predictive coding is more consistent and accurate than even manual review by lawyers.  The system has further been lauded for substantially reducing the risk of parties being accused of deliberately concealing relevant documents, as it is far easier to justify the non-production of an important document where the predictive coding programme coded it as non-responsive in the first instance.

For all its benefits, however, this new technology is not without its flaws.  The primary concern is that it is used in a small number of cases, and as such, industry standards are not set, and there is little agreement about when and how it should be used. Provided agreement is not reached between the parties to the Quinn litigation in the interim, the hearing in October will be an eye-opener as to how receptive or indeed responsive the Irish courts are to this development.

The Court could take the view that as long as a party can prove it adhered to its discovery obligations, the methodology employed does not become relevant until post production if issues arise in relation to the adequacy of discovery. As predictive coding relies so heavily on results gleaned from an initial set of reviewed documents, the comprehensiveness of this preliminary step in the process will therefore be crucial in defending any adequacy based challenge. Alternatively, and in line with the requirement in the US for an agreed discovery protocol, the Court could approve technology assisted review on the condition that all parties agree and input into its use.