For more than two decades, the courts have struggled to meet the challenges posed by an ever-increasing volume of electronic documents. English litigation has long embraced a “cards on the table” approach, under which parties have to disclose documents relevant to the dispute so that the court can decide the issues from a position of knowledge, rather than ignorance. But in an age where a major dispute may throw up hundreds of thousands, if not millions, of potentially relevant documents, such an approach can lead to huge costs unless some way is found to control the process.
In seeking to reduce the costs of disclosure, the direction of travel has (understandably) been to try to limit the amount of documentation that must be reviewed and disclosed. In general terms, that was the aim of the changes to disclosure implemented by the Woolf reforms in the late 1990s, the Jackson reforms almost ten years ago, and the disclosure pilot which began in 2019 and has recently been adopted as a permanent practice direction for the Business and Property Courts (at PD 57AD).
But the court’s decision in Genius Sports Technologies Ltd v Soft Construct (Malta) Ltd bucks that trend. In that case the court ordered the “massive overdisclosure” of documents, taking the view that this would both improve the process and save costs as compared to a more traditional approach. This post considers the merits and practicalities of such an approach and concludes that, in most cases, the downsides are likely to outweigh any potential benefits.
Successive attempts to rein in disclosure
As far back as 1882, the so-called “Peruvian Guano” approach to disclosure (or discovery) was established in a case bearing that name. Under that approach, the parties were to disclose all documents relating to any matter in question in the action, including not only those which might advance their opponent’s case or damage their own, but documents which might fairly lead to a “train of inquiry” with either of those consequences.
That approach seemed serviceable enough for the next hundred years or so, until modern business practices (including, significantly, the invention of the photocopier) led to greatly increasing volumes of documents. In his Access to Justice report in 1996, Lord Woolf recommended a move to what became known as “standard disclosure” under the Civil Procedure Rules, namely the disclosure of documents which support or adversely affect any party’s case, rather than broader “train of inquiry” disclosure.
In fact, Lord Woolf’s recommendations went further. He recognised that it was no use limiting the categories of document a party had to disclose if it still had to search through all its documents to identify them. He therefore proposed that each party would have to disclose, by way of standard disclosure, only those documents of which it was “aware”, having made enquiries of relevant individuals. But that appears to have been seen as a step too far, as there was no such limitation to standard disclosure as introduced in the CPR.
The next major change was the move to the disclosure “menu”, implemented in 2013 as a result of recommendations in Lord Justice Jackson’s Final Report following his Review of Civil Litigation Costs in 2009. This did not do away with standard disclosure, but encouraged the parties, and the court, to adopt a more tailored approach based on a “menu” of options, including an order that a party disclose the documents on which it relies while requesting any specific disclosure it requires from any other party, and disclosure of documents on an issue-by-issue basis.
In subsequent years, for the most part, the disclosure menu remained largely unsampled, with parties generally continuing to default to standard disclosure in most cases. As a result, continuing concerns at the excessive costs, scale and complexity of disclosure led to the introduction of the disclosure pilot in 2019. The disclosure menu was replaced by a new (if similar) list of “models”, but with clear signs steering the parties toward a more restrained approach to disclosure by reference to an agreed list of the key issues in dispute.
Genius Sports: massive overdisclosure
In Genius Sports, as noted above, Mr Justice Marcus Smith ordered a bespoke regime involving the “massive overdisclosure” of documents between the parties. Under this approach, the parties would not attempt to identify and produce relevant documents, in the usual way, but would instead only exclude “unequivocally irrelevant and privileged documents”, allowing the receiving party to review the rest itself.
The claimants argued that this approach went against Nichia Corporation v Argos Ltd, in which the Court of Appeal emphasised that under “standard disclosure” the disclosing party should consider each document to see whether it supports or adversely affects any party’s case, and not just disclose a mass of background documents which don’t really take the case one way or another. The court in that case said there was a “real vice” in the overdisclosure of documents, as it: (i) forced the opponent’s lawyers to read masses of documents, which were then imported into the rest of the case including the trial bundles; and (ii) led to a real risk that the really important documents would get overlooked.
Marcus Smith J commented that Nichia Corporation showed concerns about the costs of disclosure had been prevalent for many years and, despite many sets of reforms, had continued unabated. However, things had moved on since 2007, including because the “dumping” of documents no longer compelled the receiving party’s lawyers to read large volumes of material or led to the risk of important documents getting overlooked. That was, in particular, because the receiving party’s lawyers would have their own processes of electronic review to conduct targeted searches for what is relevant.
Instead, the judge said, there may be a real risk of a filtering exercise by the disclosing party leading to relevant documents being excluded so that they are never subjected to an “eyeball” review. He concluded that the massive overdisclosure approach ought to be adopted, provided that four conditions were satisfied:
- A real risk that, if a standard process of disclosure were adopted, relevant documents would be missed;
- No danger of the process being used to oppress any party, most obviously because it does not have the means to review significant electronic disclosure;
- The risk of inadvertent disclosure of privileged material is contained; and
- Confidential material is appropriately protected.
The judge concluded that these conditions were satisfied in the case in question and so the “massive overdisclosure” approach should be adopted.
Bucking the trend
As is obvious from the above, the Genius Sports decision goes against the direction of travel of successive attempts to reform the disclosure process, which have almost exclusively been aimed at the parties disclosing less rather than more – though Jackson LJ’s “menu” did include “train of inquiry” disclosure, which he said is sometimes appropriate in fraud cases.
And the “massive overdisclosure” approach is not dissimilar to a possible order that Jackson LJ recognised could be made under the “any other order” option from his disclosure menu, in a lecture in November 2011 (his seventh lecture in the Cost Review implementation programme). He described this as the “key to the warehouse” approach, under which each side would simply hand over all its documents (after removing privileged documents), so that each party could devote its resources to selecting what it regarded as helpful from other side’s documents. But although Jackson LJ said he was aware of a recent case in which a “key to the warehouse” order was made by the Technology and Construction Court, I am not aware of another reported case in which a similar approach was adopted – until now.
Merits and practicalities
The question is whether the massive overdisclosure approach has the potential to lead to an improvement in the disclosure process and/or a saving of costs – and if so in what cases.
As noted above, the order in this case stemmed crucially from a concern that the usual disclosure process would lead to a real risk of relevant documents being excluded from review – as encapsulated in the judge’s first of his four conditions listed above. But this risk is likely to be inherent to some extent in any electronic filtering process, which (as the judge recognised) is essential to any major disclosure exercise as a human review of everything would be disproportionate – and in many cases entirely impracticable or even impossible.
The decision suggests that, in the judge’s view, this risk is likely to be greater in competition cases (as this case was). As the judge put it, the issues for disclosure tend to be more “diffuse” and so it’s harder to come up with filtering techniques that can robustly identify responsive documents. But even if that is true for some competition cases, is massive overdisclosure the answer? It seems to me there are a number of significant practical difficulties which mean that this approach is unlikely, in most cases, either to improve the process or to save costs.
First, there are real conceptual difficulties as to how a filtering exercise can be conducted to exclude irrelevant documents, rather than identify relevant ones. On the assumption that the excluded documents are not themselves to be subjected to an eyeball review – which would presumably render any costs savings illusory – it seems likely that the risk of excluding relevant material would remain.
Second, even if it is the receiving party reviewing the corpus of documents, it cannot review everything and will have to use electronic search and filtering techniques to narrow down what is looked at. So, the risk of missing relevant material is not overcome, merely transferred to the receiving party – albeit that it will have a greater incentive to spot documents that are unhelpful to the disclosing party than if the disclosing party were reviewing its own documents.
Third, even if the disclosing party is not required to conduct an eyeball review of any of the documents it is disclosing, it will presumably wish to know what they contain, both to prepare its own case and be ready to defend against the opponent’s arguments. So, it will do its own searches over those documents with a view to identifying potentially relevant documents and reviewing them as appropriate – again, reducing or potentially eliminating any costs savings.
Fourth, while the decision notes that inadvertent disclosure of privileged material is a risk whatever disclosure process is adopted, that risk must be significantly greater where large volumes of documents are to be disclosed without being subject to human review. Presumably some form of privilege clawback agreement could be adopted, so that there is no waiver if privileged documents are disclosed in error (even where the error is not obvious), there is still the risk of privileged material becoming known to the opponent – and of course, once known, it cannot be forgotten.
And finally, in this case the judge recognised that there would need to be processes in place to protect confidentiality – given that the exercise is likely to result in the disclosure of large numbers of confidential documents, many of which may be of marginal relevance. These processes include strict controls on who can access the disclosed material and a requirement for individual undertakings to be given not to use the documents for any collateral purpose. And, for specific documents – which presumably will be identified by the disclosing party, but that is not clear from the judgment (nor is it clear how such documents could reliably be identified if the disclosing party has not reviewed all or most of its own disclosure) – access will have to be approved by a KC retained by the receiving party. The judge accepted that this was quite an intrusive and expensive obligation, which he regarded as somewhat exceptional. It seems that, as well as considering whether the relevant individuals really need access, the KC will be expected to assess whether the undertaking not to use for collateral purposes would be upheld by those individuals – though it is difficult to see how a KC would conduct that assessment in practice.
Overall, therefore, while there may be cases in which such an approach is appropriate, it is likely to throw up a host of practical issues which will need to be ironed out in any given case, and may make any perceived benefits illusory.
