Matthew Nelson of Symantec has an interesting article on Forbes.com this week. Called Federal Judges Consider Important Issues That Could Shape the Future of Predictive Coding Technology, it considers the possible impact of two cases presently before the US courts. Don’t look away, you UK litigators – the principles apply to you as well.
One is Da Silva Moore v Publicis Groupe et al., where the parties agreed to use predictive coding technology but have been unable to agree how it should be used. I have mentioned this case before, but am keeping my powder dry on it until we have the hoped-for written opinion from US Magistrate Judge Andrew Peck.
The second case is one I had not come across, Kleen Products LLC v Packaging Corporation of America, et al, where the discovery issues are being heard by US Magistrate Judge Nan Nolan. Here, one party seeks an order requiring the other to use predictive coding. The usual form of these disputes (going wider than any particular technology) is that Party A chooses a technology and Party B challenges its use, with the dispute generally coming down to how the technology was used rather than the particular choice of application. Now we have Party A seeking to compel Party B to adopt a particular type of application.
Matthew Nelson touches on a number of interesting points. One might be diverted by the plaintiff’s suggestion that predictive coding is akin to a car where the defendants want merely to use “the best available horse”. A couple of well-chosen Craig Ball quotations are brought in, one on court validation showing why I don’t seek to compete with Craig for colourful imagery (even I have never managed “court validation” and “steaming pile of crap” in the same sentence), and one asserting correctly that “the integrity of the process hinges on the carpenter, not on the hammer.”
The central point, one which it is vital to carry home to lawyers and judges alike, and in any jurisdiction, is that “predictive coding is one of many tools that should be included in the litigator’s tool belt”.
There is a rather fatuous argument which comes up occasionally in discussions on the UK rules – I have seen it today as it happens – as to whether the word “should” means something less than “must” in any practical sense. When Matthew Nelson says that predictive coding “should be included in the litigator’s tool belt”, he means that it must be considered in any sizeable case. The litigators must know about it and understand its capabilities, if only to be able to take part in argument about it in cases like that which are the subject of Matthew Nelson’s article.