Technology Assisted Review - The Future of Large-Scale Discovery in Ireland?

PUBLISHED: 23rd June 2015

Photo to illustrate article

A Combination of Man and Machine

 A recent decision of the Commercial Court is likely to have a dramatic impact on the way in which we manage large-scale discovery writes Amy Bradley.

In an important decision in Irish Bank Resolution Corporation Ltd & ors v Quinn & ors [2015] IEHC 175, the Commercial Court recently approved the use of Technology Assisted Review (TAR) using predictive coding in discovery. 

TAR has already been endorsed in the United States, but this is the first time that an Irish Court has been asked to determine whether a discovery using TAR will comply with discovery requirements as set out in the Rules of the Superior Courts.

It is generally recognised that discovery review in large-scale litigation can lead to significant and often disproportionate cost for parties to litigation in Ireland.  A major increase in the volume of electronically stored information has led to a significant increase in the amount of relevant data in almost every case.  Depending on the case, fees for discovery can equate to as much as 50% of the total costs in a given dispute.

In the Quinn case, IBRC's initial scoping exercise involving key word searches yielded 1.7 million potentially relevant documents.  Following the removal of duplicates and documents in other languages, that number reduced to approximately 680,000 documents.  IBRC estimated that using a traditional review, without the assistance of available technology, a team of 10 reviewers, would take 9 months, at a cost of €2 million to review the documents.  IBRC put it to the court that using predictive coding, it could make discovery within a much shorter timeframe at substantially lower cost.

How does TAR work?

In a traditional discovery, a party agrees, or is ordered by Court, to provide to the other side all data within its possession, power or procurement that is relevant to specific categories   relating to matters in dispute between the parties.  Generally, the lawyers conduct a manual review of all potentially relevant data provided by their client and they assess what is, and is not, required to be disclosed to the opposing party.

With TAR, a software programme is trained by subject matter experts, the Senior Lawyers, when they have received the potentially relevant data from their client.  The experts inform the software as to the relevance or otherwise of a small number of documents taken from the overall dataset of potentially relevant documents.  Once the software has been trained, it uses predictive algorithms and analytics to determine the relevance or otherwise of each document in the entire dataset.  This saves on the time and cost of having lawyers trawl through each and every document to assess them for relevance, as would occur in a traditional manual review. 

Numerous quality control checks are carried out throughout the process.  

What impact should this technology have?

The Commercial Court found that TAR allows for a more expeditious and economical discovery process.  The Court also found, based on the evidence before it, that discovery using TAR would probably yield more accurate results than a manual review to identify relevant documents. 

The Court’s approval of TAR is contingent upon the appropriate process being followed and also that there is sufficient transparency of process.  It is clear that any party proposing to use TAR should liaise with the other parties to the proceedings about this, as soon as possible.  Transparency is key.


The Irish judiciary appear keen to keep discovery costs as low as practicable.  The Judgment in the IBRC case does seem to represent a positive development for parties to large-scale litigation and their legal teams alike.  It is clear however that lawyers, and possibly technical advisors, should be involved from the outset of the discovery process right through to its conclusion to try to ensure the integrity of the process.   

By using this website you allow us to place cookies on your computer. Our cookies do not personally identify you.