The Epiq Angle

UK Court Says Yes to Predictive Coding in eDisclosure

4 May 2016

By Brandon Mack
Director of Advanced Technologies

As eDisclosure predictive coding continues to gain popularity in the United States, it also appears to be getting wider acceptance abroad as well—especially after a groundbreaking go-ahead from a high-level U.K. court.

Courts are always open to methodologies that help speed up the process of litigation, and plaintiffs and defendants are always looking to cut costs. Enter predictive coding, which starts with automatically pulling a representative sample from a large number of documents. After this sample is reviewed for relevance and the documents are categorized accordingly, the revised parameters are programmed into software that can categorize the rest of the documents the same way.

Pyrrho Investments Case Leads to UK Predictive Coding

In February, the England and Wales High Court, Chancery Division approved the use of predictive coding to satisfy disclosure requests (Pyrrho Invs. Ltd. v. MWB Prop. Ltd.). Although the court acknowledged that the use of predictive coding may not be a good fit for every case, it decided that—where it is possible and appropriate—this method can and should be used alongside manual review to fulfill disclosure requests for electronically stored information.

The case, brought by Pyrrho Investments, claimed that there was a breach of fiduciary duty by MWB Property. Pyrrho sought to review documents held by MWB that would help bolster its claim. However, MWB had over 17 million documents (3.1 million after deduping) to review in order to find those relevant to Pyrrho’s disclosure request. MWB sought to use predictive coding to help recognize which documents in its possession would answer the request for disclosure, as a manual search would be costly and time consuming.

Scope and Quality Important to eDisclosure

The court stated the most important part of the eDisclosure process is “the scope and quality of the search.” With over 3 million possibly relevant documents in its possession, MWB had a daunting task ahead of them. Thankfully, the availability and use of predictive coding allowed them to efficiently and swiftly complete the task. As there were few written rules on how a search for electronically filed documents must be conducted, the court found that the use of predictive coding was enough to reach the “reasonable search” standard used in disclosure requests.

After all, the court figured, less efficient methods like keyword searching have long been deemed appropriate.

In Pyrrho, both parties agreed on the method to be used in answering the disclosure request for electronic documents; the court needed only to approve of the method. The court explained that it accepted predictive coding in this case because the method:

  • would cut costs and save time,
  • was accepted by both parties,
  • had passed court muster in other countries, such as the U.S. and Ireland, and
  • could always be dropped in favor of another disclosure method if it didn’t work out, given that the trial itself is still more than a year away.
Brandon Mack

With more than 10 years of experience in the legal industry, Brandon Mack, director of advanced technologies at Epiq, is an expert in document reviews for litigation, acquisition and investigation matters. He creates and manages quality control processes for use in document review to prove the defensibility of review results used in litigation matters. For more information, contact Brandon directly at or visit the Epiq website for more information.

Filed under: edisclosure


Blog postsRSS