One of the tasks with which lawyers are faced in the electronic age is how best to deal with the sheer volume of potential documents that could be the subject of a disclosure exercise. One of the potential solutions – predictive coding – has been the subject of much debate but (in this jurisdiction at least) lacked judicial comment. A recent judgment has changed that.
The process of predictive coding requires a sample of documents to be reviewed by a senior lawyer familiar with the case to decide on those documents’ relevance. Intelligent software then creates a code which mimics the decisions made by the lawyer. The effect is that an entire set of documents can be technologically analysed on the same basis as the sample data. The results of this exercise are then reviewed by the lawyer for quality control purposes to check that the predictive code is flagging the correct set of documents. If an irrelevant document has been incorrectly flagged as relevant (or vice-versa), the lawyer will amend this, and that change will be picked up by the software and the decision applied across the whole body of documents. This process is repeated until the lawyers are comfortable that the exercise has generated a set appropriate for manual review. As well as providing a consistent initial review, by reducing the document set for manual review predictive coding can turn what could otherwise be an exorbitant exercise into a cost-effective and proportionate one.
In Pyrrho Investments Limited and another v MWB Property Limited and others  EWHC 256 (Ch) (16 February 2016), these and other considerations were explored. This multimillion pound case involved the potential review of 3.1 million electronic documents and the parties, subject to the court’s approval, had agreed amongst themselves that predictive coding may be beneficial. The Court agreed, finding that there were no factors of any weight against the use of predictive coding.
Although predictive coding will not be appropriate for every case, this decision is welcomed as providing judicial approval for one potential method of reducing the potential efficiency, time and cost of disclosure exercises.