Predictive coding finally receives legal approval as an authoritative method in E-disclosure in breakthrough ruling
A landmark case in Ireland has finally cemented the protocol that predictive code technology will be accepted in the disclosure process. The case, Irish Bank Resolution Corporation Ltd & ors v Quinn & ors  IEHC 175, quashes any concerns certain parties may have voiced in the past about its accuracy, safety or relevance. The bottom line is that by litigants becoming more confident in using advanced analytical techniques in litigation disclosure, it can save time and money.
By nature disclosure tends to be expensive, e.g. in straightforward cases where there’s one claimant and one defendant there can potentially be documents running into the thousands or hundreds of thousands whilst more complex cases with higher numbers of claimants going after multiple defendants (common in commercial cases) the number of documents can run into the millions.
Predictive coding is a process which involves the review of documents using computer algorithms to return expected documents based on the selection of relevant documents by a human subject expert. By using algorithms, any increase in the number of documents does not necessarily increase costs, when compared with a manual search.
For the skeptics, it has to be mentioned that it is certainly not a free for all, it has to comply with a strict methodology, e.g. both parties agree the parameters such as data size, margin of error and criteria for inclusion of documents like date range and keywords; an illustrative samples has to be run through the software to train it and checked by a human. Following that, the sample is analysed by the predictive coding software for common language and concepts and grants it a score; it is then double checked with further computer statistical sampling for quality assurance, and so on and so forth.
The use of this technology in discovery has been in use in the US for some time. In fact, the Ireland ruling went one step further by governing that an e-discovery process would not require a manual document review to be carried out, and even alluded in the discovery of large data sets, technology-assisted review (TAR) using predictive coding could even be more accurate than a manual review.
The ruling was comprehensive and was not afraid to answer certain misgivings such as whether predictive coding could miss relevant documents? The judge answered that relevant documents will always be missed just like in a human review, but as long as the correct validation steps were taken and the predictive coding relevance scores were set high enough then the obligation would have been met.
Of particular interest to English lawyers, the ruling judge stated that we should always keep in mind issues of access-to-justice when considering the all-important proportionality in Civil Procedure Rules and predictive technology was thus given its blessing in terms of the benefits of employing it.
There will never be a perfect system in place but we are certainly one step closer to a pragmatic and streamlined future.