During the past several years, discussions and commentary pertaining to predictive coding have proven more theoretical than practical. Estimates suggest that companies and law firms are utilizing predictive coding technology on less than five percent of their document reviews (based on our experience, five percent may be generous).

Predictive Coding and DiscoverReady

In his last blog on predictive coding, DiscoverReady CEO Jim Wagner asserted that 2012 might well be the year that predictive coding (or automated review or computer-assisted review or technology-assisted review or whatever you care to call it) finally gains a foothold in the document review space. Based on a recent hearing in front of Judge Andrew Peck in the S.D.N.Y, it looks like Jim’s prediction is right on target.

Judge Peck, E-Discovery Judges in Charlotte

First a little background. During the past two months, the team here at DiscoverReady has been fortunate to interact with Judge Peck on several occasions. In January, Judge Peck travelled to Charlotte to participate in a panel discussion aptly titled “E-Discovery Judges in Charlotte.” The discussion was structured as a mock 26(f) conference/Rule 16 hearing in which Judge Peck joined fellow magistrates Grimm (D.Md.), Facciola ( D.D.C.), Keasler (W.D.N.C), and Hodges (D.S.C.) in providing commentary on e-discovery based on hypothetical cases enacted by representatives from Charlotte corporate heavyweights Bank of America, TIAA Cref, Duke Energy, and Rack Room Shoes, e-discovery counsel from leading law firms, and e-discovery consultants. During this wide-ranging program, Judges Peck echoed his frequently cited sentiments — most recently expressed in “Search, Forward”— that at least certain members of the judiciary would favorably view a discovery process that utilized a documented, defensible predictive coding methodology.

We next ran into Judge Peck at LegalTech, where, during another panel discussion, he again promoted the potential efficacy of computer-assisted review while questioning the efficacy of keyword-search-based discovery methodologies. 

Judge Peck and Da Silva Moore v Publicis Group, et al

Against that backdrop, Judge Peck took the leap that many of the participants in the two panels have been anxiously anticipating. As first reported by LTN, during a February 8 discovery conference in the Da Silva Moore v Publicis Group, et al., matter, Judge Peck entertained a joint discovery proposal wherein the parties proposed the use of predictive coding technology.  Amid the subsequent buzz, many have erroneously suggested that Judge Peck “ordered” the parties to adopt a protocol for e-discovery  that includes the use of Recommind’s predictive coding technology. To the contrary, Judge Peck actually approved a protocol where the parties proposed utilizing an automated solution but differed over the specific manner in which it would be implemented.

Da Silva Moore v Publicis Group Hearing Transcript

The transcript of the hearing provides several important considerations for parties who are considering incorporating automated technology into their discovery plan:

• First, technologists can and should play a critical role in developing and validating the proposed automated strategy. In this case, technologists, rather than the counsel of record, were primarily responsible for explaining the proposed disputes and solutions to Judge Peck
• Second, these technologists have to be able to convey complex subject matter (sampling, precision, recall, confidence score) in an easily understood manner. Judge Peck likely has as much, if not more knowledge, regarding automated review than any of his peers, yet on several occasions he had to ask the technologists to “dumb it down”
• Third, automated review always will be dependent on the decisions that humans make with respect to the documents. As the parties noted and Judge Peck agreed, a “system is only as good as the training that it gets”
• Fourth, like manual review, there still will be errors associated with automated reviews.  As Judge Peck noted, automated review technologies “certainly work better than most of the alternatives, if not all of the alternatives. So the idea is not to make this perfect; it’s not going to be perfect. The idea is to make it significantly better than the alternative without nearly as much cost.”
• Finally, as we have suggested, the age of automated review may change the way parties approach and conduct discovery. In this instance, Judge Peck expressly noted that this was not a “black box” solution, since the plaintiffs were “going to see all the documents used to train the system” so they would have complete knowledge as to how the system was trained to find relevance. 

Here at DiscoverReady, we’ve spent years thinking about these and other issues, advising clients on how to conduct an effective automated review process using our proprietary i-Decision process or third-party technologies. We join the rest of the e-discovery world in following this case closely, particularly if we reach the long-awaited holy grail of a formal order approving the use of automated review technology, specifically predictive coding, to conduct discovery.

In the interim, there are many other interesting issues addressed in the full transcript. So take a look while we wait to see how this turns out. It may just be a glimpse into the future.

The field of legal technology, and e-discovery in particular, is interesting. There’s a sentiment that the “favorite technology” changes every three to five years. I think that’s mostly true. Then there’s the saying that when it comes to this industry, the faces never change. Besides a handful of wrinkles, I fear that I’m one of those faces. And then there’s the truism that the legal industry is painfully slow to change. Don’t hold your breath, but I think, on this count, 2012 will be different.

The overwhelming noise that you hear — from clients, from courts and from would-be solutions providers (other than staffing companies, of course) — is the consensus call for predictive coding.

Traditional Document Review

There is no argument that traditional document review is labor intensive and that the results are inconsistent. There is no argument that, for most, the use of keyword searching is an inexact science. And one cannot dispute that the cost and inefficiency of discovery often make litigating a case a daunting prospect at best. So, at long last, we’re going to have predictive coding. And now we ask the dog about to catch the bumper, “How do you think that’s going to taste?”

Predictive Coding

Amid the raucous cheerleading for predictive coding, we still must ask some fundamental questions:

• Does this mean that litigants finally will be required to disclose their discovery practices?
• How good does a predictive-coding tool have to be to be defensible? If manual review is not much better than the flip of a coin, which some commentators and studies have suggested (erroneously I believe), then is a predictive-coding tool that’s right more than half the time good enough?
• We know that the leading legal commentators want sound statistical sampling. But what does this mean? Who is going to design the protocols? Who is going to execute them? Measure the results? Document the process?
• Most importantly, will predictive coding actually save clients money? If a tool requires hundreds of senior-attorney training hours before spitting out large blobs of data that it “thinks” are relevant — which then require senior-attorney time to review before production — is this really saving money?

DiscoverReady Predictive Coding

DiscoverReady began addressing these topics in 2007, when we developed our i-Decision® automated review process. Since that time, we’ve refined our automated review solutions, working for some of the world’s largest companies, to provide the most comprehensive, transparent and defensible predictive coding process in the industry. Our process does not further burden senior counsel. And, most importantly, our clients don’t have to guess whether they are saving money.

So, welcome to 2012, predictive coding. We’ve been expecting you.

With its recent revisions to Utah Rule of Civil Procedure 26, Utah joined a growing number of entities focused on driving reasonableness and proportionality to the discovery process. Unlike its recent predecessors, the Federal Circuit Advisory Committee’s Model Order and the District of Delaware’s Default Standards, Utah’s proposed revisions to URCP 26(b) are not limited to the e-discovery context. Instead, as explained in language of the rule itself and the accompanying committee notes, Utah proposes a more fundamental shift for the entire discovery process, whereby discovery is conducted under a “proportionality standard.”   

In Utah’s view, this simply means that the “cost of discovery should be proportional to what is at stake in the litigation.” This certainly is not an earth-shattering proposition — as the committee notes acknowledge, the concept of proportionality is hardly new.  However, while the prior version of URCP 26 (as well as its federal counterpart) provides mechanisms to address disproportionate discovery, they were rarely invoked under either the Utah rules or federal rules. Against this backdrop, URCP 26 dictates that proportionality will become the guiding standard for discovery in all cases through several key provisions:

• First, the amendments seek to reduce discovery costs in all matters by imposing additional early disclosure requirements. These requirements dictate that parties provide critical information before general discovery begins, including
  • the production of all documents the party may offer in its case in chief and all documents referred to in the pleadings,
  • the identification of each fact witness the party may call and
  • a summary of each witness’ anticipated testimony.
  • Following the exchange of this information, which is intended to “automatically permit each party to learn the witnesses and evidence the opposing party will offer in its case in chief,” Utah envisions that all subsequent discovery will serve the more limited function of identifying the weakness in the producing party’s case. Consequently, Utah courts will limit the volume and length of discovery based on the amount in dispute, as shown below:

• To obtain discovery beyond those provided in the rules, a party must
  • reach the limits of standard discovery imposed by the rule prior to the close of discovery (preventing preemptive claims that the requirements of a particular case necessitate more discovery) and either
  • file a stipulated statement that extraordinary discovery is necessary and proportional and that each party has reviewed and approved a discovery budget or
  • file a motion setting forth the reasons why extraordinary discovery is necessary, certifying that the party has reviewed and approved a discovery budget, and certifying that the party has satisfied its obligation to meeting in an attempt to achieve a stipulation.  
  • Should discovery disputes arise, the party seeking discovery — rather than the responding party — always has the burden of showing proportionality and relevance.
  • That being said, if the responding party contends that certain electronically stored information is not reasonably accessible, the responding party must describe the source of the electronically stored information, the nature and extent of the burden, the nature of information not provided, and any other information that will allow the requesting party to evaluate the claim of undue burden or expense.

This shift to a “proportionality”-based discovery focus could have long-lasting implications in Utah while serving as a model to other jurisdictions. At a minimum, parties will be forced to establish the bounds of proportional discovery prior to engaging in broad fishing expeditions under the guise of identifying all relevant of potentially admissible information.

It will be interesting (or as interesting as e-discovery can be) to see how this plays out in practice. Don’t be surprised to see Utah’s influence loom large as federal and state courts continue to focus on driving reasonableness and proportionality in the discovery process.  

Recognizing the merits of the Federal Circuit’s Model Order — about which I have previously blogged here and here — Delaware’s Default Standard for Discovery, Including Discovery of Electronically Stored Information may prove to be an even more effective tool in encouraging parties to cooperatively resolve disputes.

As its name suggests, the standards are built upon the critical presumptions that “parties are expected to reach agreements cooperatively on how to conduct discovery” and that “parties are expected to use reasonable, good faith and proportional efforts to preserve, identify and produce relevant information.” As such, the Default Standards will only apply when the parties fail in these efforts, leaving them “unable to agree on the parameters and timing” — recognizing that even the Default Standards can be displaced by a further order of the court or subsequent agreement of the parties. 

While fairly simple in practice, the Default Standards ultimately may mark a fundamental shift in how parties address their Rule 26 “meet-and-confer” obligations. The 2006 FRCP Amendments implicitly presumed that parties would approach the meet-and-confer process in a cooperative and reasonable manner. To that end, here at DiscoverReady, we’ve spent the last five years advocating the benefits of reasonable, appropriate discovery limitations. But far too often we’ve seen discovery negotiations devolve to a point where each party tries to inflict maximum cost and burden on the opposing party.    

The Default Standards seek to remedy this problem by providing an immediate, well-defined consequence if parties fail to meet their meet-and-confer obligations. Specifically, in the absence of an agreement between the parties, the Default Standards will:

• Establish the parameters of parties’ duty to preserve discoverable information and define the categories of ESI that need not be preserved
• Require parties to define the scope of their privilege logging obligations and reaching an non-waiver order pursuant to FRCP Rule 502 (recognizing that until such order is entered, privileged information will be immediately returned if it appears to have been inadvertently produced or notice is provided within 30 days of the production)
• Require that the parties identify 10 custodians most likely to have discoverable information (at least implicitly suggesting that email review and production will be limited to these custodians)
• Require that the parties exchange a list of non-custodial data sources likely to contain non-duplicative information
• Require that producing parties disclose any search terms used to identify potentially responsive data while limiting the  requesting party to no more than 10 additional search terms absent a showing of good cause 
• Establish a standard PDF or TIFF production format, while allowing only files that are not easily converted to image format (such as Excel) to be produced natively and
• Establish the metadata fields the parties are obligated to provide.

While the Default Standards also adopt provisions for patent cases that provide a timeline for infringement and invalidity disclosures (bringing Delaware in line with other jurisdictions utilizing local patent rules) and default source code standards, the limitations above likely will have the greatest impact on the manner in which litigants — as well as courts in other jurisdictions — approach discovery. 

Other courts are going to be watching the District of Delaware very closely to see if the Default Standards work. If they do — and they should, since parties will have a clear understanding of the discovery limitations they will face if they cannot reach agreement on a discovery plan — we are likely to see more jurisdictions follow suit.  

This position is expected to provide strategic direction and have execution accountability for global human resources at DiscoverReady.   The Human Resources Director will be accountable for all human resources activity at all DiscoverReady locations, and be a strategic voice to senior leadership as DiscoverReady expands in the US and globally. The HR Director will be responsible for integrating closely with the Dolan HR team, and implementing the standardized processes to support the continued growth of the company in size and complexity.

Primary Responsibilities

• Developing and executing to the strategic roadmap for DiscoverReady. This includes defining and implementing strategic HR initiatives. 
• Supporting senior leadership in assessing and supporting entry into new markets from a human resources perspective.
• Develop staffing strategies and implementation plans and programs to identify talent within and outside the company.  Identify and manage appropriate and effective external sources for candidates for all levels within the company.  This includes recruitment for regular and contracted labor.
• Evolving the performance management program, implementing additional processes as necessary to ensure execution to standard performance review processes, but also ensure that DiscoverReady has an active talent and succession strategy in place for key roles.
• Developing a learning and innovation strategy to ensure that we are efficiently and effectively training our employees in necessary skills and industry acumen. The HR Director will also be responsible for developing processes to inspire a culture of innovation at DiscoverReady.
• Triaging employee relations issues, informing and consulting DiscoverReady senior leadership and Dolan HR when appropriate.
• Developing and managing adherence to DiscoverReady policies, including Dolan policies.
• Responsible for successful execution of transactional HR activities (payroll, recruitment, on boarding, etc.), managing a HR Specialist who will be responsible for the day to day execution of these activities.
• Maintains knowledge of industry trends and employment legislation and insurance company compliance.

Requirements

• Six to ten years related experience.
• Strong organization skills with a keen ability to prioritize and multi-task.
• Ability to adhere to and meet deadlines.
• High level of interpersonal skills to handle sensitive and confidential situations and documentation.
• Excellent communication skills (oral, written and presentation).
• Strong problem solving skills, with the ability to gather and analyze information skillfully.
• Strong knowledge of all pertinent federal and state regulations and compliance requirements.  This includes FMLA, ADA and DOL requirements. 
• Works effectively and relates well with others.  The employee exhibits a professional manner in dealing with others and works to maintain quality working relationships.
• Ability to travel as necessary.

Macs in Corporations

Historically, the use of Macs in the corporate enterprise was occasional, and limited to small groups. Most of us think of the creative types in marketing and graphic design as the most common examples of Mac users. 

But use of Macs is no longer so rare. CTOs are relaxing their policies to allow the use of Macs (which use the Mac OS X operating system), as well as iPhones and iPads (which use the iOS operating system). More and more employees – sales teams and engineering groups in particular – have embraced this flexibility and increasingly choose Macs as the working platform of choice. C-level employees have become especially high users of Mac and iOS devices. This growing corporate use of Apple devices has major ramifications in the e-discovery context, where data from these devices must be preserved, collected, processed, reviewed, and produced.

Mac Data and E-Discovery on Macs

For years, Apple’s marketing campaigns revolved around the simple slogan “Think Different.”  When it comes to dealing with e-discovery from custodians using Macs, it is wise to take that slogan to heart.  Mac data is inherently different from Windows data. Mac data uses a different file system (Mac uses HFS and HFS+ while Windows uses FAT / NTFS) and different file types and structures. 

Mac vs. Windows Data Difference

One significant example of a Mac v. Windows data difference is the “bundled” file format, a file type unique to Macs that is used by certain (although not all) versions of Apple applications such as Keynote. When loaded into a Windows environment, the system awkwardly interprets a bundled file as a series of files and folders; in a Mac environment, the system properly interprets the file as a single document. Complicating matters further is the flexibility Mac hardware offers; users can run Mac OS X or Windows – or both concurrently – on their Mac hardware. Even the files generated in Windows-based programs, such as Microsoft Office applications, are stored differently when found in a Mac operating system.

Because of these differences, Mac data must be handled differently from Windows data when putting the data through collection, processing and review. It is not possible to collect, process and review Mac data using a Windows system and still maintain a true “native” review, which is considered to be best practice.  Rather, because a Windows system is forced to interpret the Mac data, the review will be non-native – and, quite often, the processed data set will be riddled with errors and omissions.

The reality, however, is that logistics and cost considerations make it difficult in most instances for e-discovery of Mac data to be conducted entirely in a Mac environment.  But even in the absence of a native Mac collection and processing system, there are some ways to mitigate the risks of working with Mac data:

1. Collect natively. Avoid copying data onto NTFS or FAT formatted drives and avoid using proprietary image formats (L01). Using these methods may alter metadata such as dates, file names, type and creator codes, data and resource forks, and file paths.  Instead, collect data using raw formats such as .dd or .dmg, and copy the data onto HFS+ formatted drives. 
2. Perform early case assessment on a Mac. This may not be possible if you did not retain the original files in a raw format on HFS+ formatted drives. But if you did, you should review the data natively in a Mac environment for your early case assessment.  Be careful, however, to take the appropriate precautions to protect the integrity of the files (locking, mounting read-only, etc.).
3. Render the Mac data in a format that retains the most accurate representation of the native data. The ideal format will maintain the critical metadata for each file, but will integrate well into Windows-based review solutions. PDF is an optimal format for Mac data in the absence of a native Mac review platform. 
4. Avoid using “standard” file-type processing lists, which typically are Windows-centric.   Files from applications such as the iWork suite (the Microsoft Office equivalent on the Mac), Omni products such as Graffle and Outliner, and other Mac-only applications may all contain important data.  If you use a “standard” file type list that does not include these files, you may inadvertently omit key data from processing.

Macs and the data found on them are different from Windows-based PCs, and they should be treated as such. As their presence grows in the enterprise, it is important for legal professionals handling e-discovery – both in the corporate in-house legal department and in law firms – to understand these differences and accurately assess how best to collect and process Mac data.  And Macs are only the beginning.  In the coming years, iPhones and iPads – with their equally unique iOS operating system – will contain a flood of potentially responsive data that should be treated with similarly nuanced understanding and care.   

Over the past several weeks, courts in the Eastern District of Texas and the Northern District of California have provided a preview of how the Model Order could play out. 

Stambler v. Atmos Energy (Case No. 2:10-CV-594)

On September 29, just two days after Judge Rader’s speech, the Eastern District of Texas Court used the Model Order to resolve a pending discovery dispute in the matter of Stambler v. Atmos Energy (Case No. 2:10-CV-594). In Stambler, Judge Everingham faced competing proposals where, among other things, the plaintiff asserted that it was entitled to email production from 20 custodians using 20 search terms to cull the data.  The defendant did not directly respond to the proposed limitations, instead suggesting that the email discovery should be delayed, and that the plaintiff should not dictate how the defendant identified and produced data. In resolving the dispute, Judge Everingham did not directly make use of the Model Order, but instead issued a short discovery order that: 

• rejected defendant’s argument that email production should be delayed
• initially limited email collection to five custodians that would be culled using 10 search terms and
• provided that either party could move to expand the number of custodians or search terms upon a showing of good cause.

Effectively Illuminated Pathways v. Aston Martin Lagonda (Case No. 6:11-CV-34)

On October 20, Judge Love in the Eastern District of Texas issued a discovery order in Effectively Illuminated Pathways v. Aston Martin Lagonda (Case No. 6:11-CV-34) that invoked at least the spirit, if not the complete substance, of the Model Order. In Effectively Illuminated, the plaintiff had proposed limiting email discovery to five custodians, while defendants proposed delaying email production until the parties could determine whether email discovery was necessary. Following a phone conference, the parties jointly entered a proposed order that:

• delayed email discovery pending the initial production of each party and
• provided that, following the initial discovery, each party could request the email production of five custodians, with the searching to be done in “the most efficient way possible with fewest and most relevant search terms.”

Intravisual v. Fujitsu (Case No. 2:10-CV-090)

On October 21, Judge Folsom in the Eastern District of Texas issued in Intravisual v. Fujitsu (Case No. 2:10-CV-090). Critically, in this case, the parties jointly submitted a proposed order that essentially tracked the Model Order — and Judge Folsom entered the order as proposed.

DCG Systems Inc. v. Checkpoint Technologies LLC (Case No. C-11-03792)

Finally, on November 2, the Northern District of California got in on the action when Judge Grewal issued a discovery order in DCG Systems Inc. v. Checkpoint Technologies LLC (Case No. C-11-03792). In DCG, the defendant proposed a discovery order adopting the Model Order, while the plaintiff opposed adoption, arguing that the court rely on the limits imposed by FRCP 26 and 34 to generically establish the boundaries of email discovery. 

While praising the parties for working diligently to “address many of the ESI issues that often plague” complex patent cases, Judge Grewal noted that the parties could not reach accommodation on the subject of the Model Order. The inability to do so was premised primarily on defendant’s position that the Model Order was most applicable to cases involving non-practicing entities, and should not be applicable in a case involving direct competitors. Judge Grewal rejected the defendant’s argument, and adopted the plaintiff’s proposed version of the Model Order, noting that:

Perhaps the restrictions of the Model Order will prove undue. In that case, the court is more than willing to entertain a request to modify the limits.  But only through experimentation of at least the modest sort urged by the Chief Judge [Rader] will courts and parties come to better understand what steps may be taken to address what to date has been a largely unchecked problem.

So what can we take away from these initial tests of the Model Order? First, as I anticipated in my last blog entry, it is unlikely that the Model Order will be applied in a uniform fashion. Instead, courts likely will enter discovery orders with reasonable limitations jointly agreed to by the parties. Conversely, we also expect courts to use the Model Order in whole or in part to impose constraints on parties that are unable to reach agreement on reasonable discovery limitations. 

Second (and perhaps obviously), if your opposing party is coming to the table with discovery limitations that are in line with the Model Order, the days of simply objecting to their proposal is long gone. Parties should immediately start considering the limitations of the Model Order when they are negotiating discovery plans with their adversaries. If you go in just saying no to limitations on discovery, you are likely going to find yourself operating under the Model Order or some close variation thereof. This likely will be the greatest effect of the Model Order, as parties will need to cooperate in developing reasonable discovery plans. 

Finally, and perhaps most interestingly, it appears that the courts, like most practitioners and commentators, are still weighing the relative benefits of the Model Order. Based on the precedent to date, courts appear reluctant to impose hard-and-fast limitations on custodians and search terms unless the parties jointly agree to them. Judge Grewal’s commentary perhaps best articulates the apparent sentiment that courts are not completely sold on all of the provisions of the Order, but are going to give it a try until they see how the custodian and search-term limitations shake out.

Amid this frenzy of activity, some of our busiest senior-level team members from both organizations recently took time from their schedules to step back and strategically address one of our industry’s most critical issues — the topic of quality in document review. 

As we’ve explored time and again, assessing quality or “accuracy” in the context of document review is a complex topic. If a document is marked responsive and one partner on the case agrees with the decision and another partner disagrees, which one is right? In those instances, is the right answer that there’s no right answer? Separately, if a document is marked as responsive and three issue tags are applied, is the document decision wrong if a fourth tag is missing? 

These are not trick questions. And the obvious answers may not be the right answers, especially as our industry is trying to sort through the “accuracy” of human review versus the accuracy of automated review (aka predictive coding).

I’m very proud that over the years our combined organizations have been on the forefront of addressing both “what constitutes quality” and “how to generate quality” in document review. And as much as DiscoverReady and ACT have done separately on this subject, I’m even more excited about the work to come from our newly unified team. 

Quality in Document Review

In the coming weeks and months, Macyl Burke, our Valencia-based guru on quality in e-discovery, and David Shub, our senior-most legal contributor to the topic of achieving quality in document review, will be working hard to provide a practical guide to defining and achieving quality in document review – human or automated. 

The first charge for these gentlemen, and a multitude of colleagues working behind the scenes, is to lay the foundation for a common understanding of “defining and assessing quality in document review” that can be accepted throughout our industry.

From an internal perspective, this team also is responsible for creating a roadmap of best practices for our combined organization to follow going forward. These practices will incorporate the best of planning, process, automation and sampling, to ensure that we continue delivering the highest levels of quality and defensibility in document review. 

Stay tuned for more. 

• Competent professionals executed the review
• Reasonable steps were taken to ensure the quality of the process
• The process was well-documented
• The process was transparent, such that the basis for any automated decision can be readily understood and explained
• The parties conducted adequate sampling to ensure accuracy rates as well as appropriate precision and recall

In an apparent case of great minds thinking alike, Magistrate Judge Peck of the Southern District recently echoed Jim’s thoughts in a Law Technology News article.

Before addressing the reasonableness and defensibility of automated review technologies, Judge Peck first skewers the notion that human review of documents is the “gold standard” of document review process. It should not be, in light of studies that “clearly demonstrate that computerized searches are at least as accurate, if not more so, than manual review.” He also examines the well-accepted use of keyword searches in connection with human review, noting that keyword searching often is fraught with problems and typically generates very low precision and recall. 

Judge Peck also shares Jim’s observation about precedent, remarking that “[t]o my knowledge, no reported case (federal or state) has ruled on the use of computer-assisted coding. While anecdotally it appears that some lawyers are using predictive coding technology, it also appears that many lawyers (and their clients) are waiting for a judicial decision approving of computer-assisted review.”

While acknowledging that such a decision may be a long time coming, Judge Peck offers guidance to litigants interested in using these technologies now. From his judicial perspective, if faced with a challenge to the use of predictive or automated coding, he will want to know “what was done and why that produced defensible results.” He would expect to see statistics on the technology’s precision and recall. “Proof of a valid ‘process,’ including quality control testing,” will also be important in establishing defensibility, he writes.

Judge Peck concludes his article by saying this:  “Until there is a judicial opinion approving (or even critiquing) the use of predictive coding, counsel will just have to rely on this article as a sign of judicial approval. In my opinion, computer-assisted coding should be used in those cases where it will help ‘secure the just, speedy, and inexpensive’ (Fed. R. Civ. P. 1) determination of cases in our e-discovery world.” 

Thanks, Judge Peck.  We couldn’t have said it better.