Judgment given in the English High Court in February marks the UK’s first judicial approval of the use of predictive coding in disclosure.

Kylie Peterson – Director of Consulting & Services

Kylie Peterson – Director of Consulting & Services

Kylie Petersen, lawyer and Director of Consulting & Services with e.law International, explains the significance of the case.

The proliferation of electronically stored information has created a range of challenges for courts and litigants. Chief amongst these is the challenge of achieving effective disclosure at a reasonable cost and within reasonable timeframes.

In 2012 in the unreported case of Moore v Publicis Groupe, US District Court Judge Andrew Peck made the first judicial pronouncement of the role of predictive coding in meeting this challenge. Acknowledging that there is no review tool – human or automated – that guarantees perfection in document review, Judge Peck recognised the appropriateness of predictive coding as a review tool. This position was endorsed by Fullam J of the Irish High Court in Irish Bank Resolution Corporation Ltd v Quinn [2015] IEHC 175.

On 16 February, in Pyrrho Investments Limited and MWB Business Exchange Limited v MWB Property Limited and others, Master Matthews of the English High Court reviewed the requirements of the Civil Procedure Rules 1988 and relevant Practice Notes. As these contained no prohibition, and on the basis of a range of factors including experience in other jurisdictions, accuracy considerations and cost and proportionality factors, he approved the use of predictive coding. This amounted to the first judicially-endorsed use of predictive coding in the UK.

What is predictive coding?

Predictive coding is also known as technology assisted review, computer assisted review or assisted review. In general terms, predictive coding means that, after an initial training period involving the creation, review and recreation of sample sets of documents, software is used to generate customised search algorithms for identifying and categorising responsive and privileged documents. The results are validated through human statistical sampling and the results are fed back into the software for further learning and categorisation. Only a small proportion of the entire document set will be subject to human review.

The Australian position

To date, there has been no judicial determination by Australian Courts of the appropriateness of predictive coding in disclosure. No reference is made to predictive coding (however described) in Civil Procedure Rules or Practice Notes although all, including Practice Note SC Gen 7, “encourage the use of information technology as a means of improving the efficiency of litigation”. Practitioners should be aware that current generation eDiscovery software incorporates predictive coding capabilities and that there is nascent but growing use of this capability by Australian litigators. For the sake of certainty, it is hoped that our Courts will evaluate the appropriateness of predictive coding in meeting disclosure obligations in Australia.

To find out more about predictive coding, call 1300 30 30 80 or email Kylie Petersen directly on [email protected]