Co-Authored by: Judy Selby and Jessica Nutt

In prior posts, we’ve commented on the emerging judicial focus on the use of technology-assisted review/predictive coding, as well as the twists and turns of Da Silva Moore, in which Magistrate Judge Peck permitted the use of predictive coding in a groundbreaking decision, only to face demands for his recusal or disqualification. 

On June 15th, in a 56 page opinion and order, Judge Peck denied plaintiffs’ motion for his recusal “as untimely and in any event as without merit.” While the merits of predictive coding were not front-and-center in the opinion, the factual background provides an interesting outline of how the parties approached the development of search and review protocols, how they worked with each other, and how they worked with the court. While the opinions in Da Silva Moore focus on disputes over the use of a nascent approach to the review of voluminous electronic data, they are also interesting in that they highlight the high level of discussion that parties should engage in when faced with significant discovery burdens. With that said, as Da Silva Moore shows, not all high level discussions end well.

A few observations on Judge Peck’s order denying plaintiffs’ motion for recusal:

  • The opinion lays bare a key troubling aspect of the recusal motion:  plaintiffs’ apparent preference that jurists remain blank slates to the point where they do not educate themselves, or participate in the education of others, concerning critical emerging litigation matters that are largely responsible for driving the significant costs of litigation. Judge Peck reasoned that participating in speaking engagements – such as eDiscovery and technology-assisted review panels – is encouraged and does not inherently show bias.  In fact, the Code of Conduct for United States Judges, Canon 4, permits judges to engage in these types of activities. 
  • The opinion focuses on the timeliness of plaintiffs’ motion, which came long after Judge Peck’s acknowledgment of his involvement in the eDiscovery world and his particular interest in predictive coding. This, coupled with plaintiffs’ vendor’s support of predictive coding, and trumpeting of its own role in predictive coding in Da Silva Moore, suggests that the motion for recusal may be more of a litigation tactic than an actual attack on predictive coding (although many would challenge the wisdom of this motion as a litigation tactic).

We expect this latest ruling will renew spirited discussion in the blogosphere and beyond about the emergence of predictive coding as a tool to cut litigation costs.  But, we remind our readers, whether we are talking about Da Silva Moore, Kleen Products, or Global Aerospace, we are waiting for the other shoe to drop. While courts are allowing the use of predictive coding, they are also recognizing that the parties may dispute the results of the process. Not a single court has rubber-stamped technology-assisted review as used in a particular case as defensible. 

The bottom line: the iterative process of technology-assisted review is likely to see further judicial iteration as well.