In a recent decision, the High Court has declined to order Model E disclosure under the disclosure pilot, and has also held that data and documents on a server held by a company were not within the control of a director and a former director of that company (who were defendants to the action) so as to fall within their disclosure obligations: Kelly v Baker [2021] EWHC 964 (Comm).

Model E or “wide search-based” disclosure, under the disclosure pilot at PD 51U, requires a party to disclose documents which are likely to support or adversely affect any party’s case in relation to one or more of the Issues for Disclosure, or which may lead to a train of inquiry which may then result in the identification of other documents for disclosure. This equates to old-style Peruvian Guano or “train of inquiry” disclosure, which was replaced by “standard disclosure” when the Civil Procedure Rules were introduced in 1999. The court has retained a discretion to order this form of disclosure, and it was one of the “menu” of disclosure options introduced in place of standard disclosure under the Jackson reforms in 2013, but it has long been understood to be available only in exceptional cases (see for example Berezovsky v Abramovich [2010] EWHC 2010, considered here). This requirement of exceptionality is stated expressly in PD 51U.

Against that background, the present decision underlines that exceptional really means exceptional. A party seeking Model E disclosure will, it seems, have to do more than show that it is an important, high value case, or that it involves allegations of fraud. Further, the disclosure sought must be linked with specific issues and some explanation provided as to the nature of the enquiry envisaged.

The decision also illustrates the fact-sensitive nature of the question of whether documents are in a party’s “control” for the purposes of disclosure. It suggests, perhaps unsurprisingly, that a company’s documents are unlikely to be considered to be in the control of a director, in their personal capacity – at least where it is not a one-man company. It is also of interest in taking a relatively strict view of whether documents held by a third party are in a litigant’s control, picking up on comments in Pipia v BGEO Group Ltd [2020] EWHC 402 (Comm) (considered here) that a relevant factor is whether the party has “free and unfettered access” to the documents in question. This may be seen as at odds with the view expressed in Berkeley Square Holdings Ltd v Lancer Property Asset Management Ltd [2021] EWHC 849 (Ch) (considered here) (which is not referred to in the present decision) that the references in Pipia to free or unfettered access were “unhelpful” and could cause confusion as to the need for a party to be able to access the documents by whatever means they see fit, which is not a prerequisite for control in the relevant sense.


The underlying proceedings relate to a management buy-out. The claimant’s case (in essence) is that he placed reliance on the defendants to get the best price on the market for the relevant assets and that they assumed a fiduciary duty to him in advising him and dealing with third parties.

At a case management conference in April 2021, the court considered a number of issues relating to disclosure including:

  • whether the defendants should be required to give “Model E” disclosure; and
  • whether the defendants had control of data on a server which had been transferred to the purchaser company as part of the transaction.


The court (Moulder J) held that the claimants had not justified Model E disclosure, and the appropriate order was Model D. She also held that the defendants did not have control of the server data.

Model E disclosure

The claimants argued that Model E was appropriate on the basis that the case involves allegations of fraud, which had been met with a bare denial, and the emails of the defendants and others who were employees of the companies purchased would be key to establishing what the defendants were telling the claimant and his family members leading up to the transaction, what they were telling the outside world about their roles, and what they were telling the professional advisers involved in the transaction. They submitted that Model E disclosure was justified by the factors set out in paragraph 6.4 of PD 51U, in particular that: there were serious allegations of dishonesty; the value of the case was high; the financial position of the parties was not a barrier to Model E disclosure; and there was one dataset. The defendants argued that Model D should be ordered.

Moulder J said the starting point, as clearly set out in PD 51U (para.8.3(2)) is that Model E disclosure is only to be ordered in an exceptional case. It was not enough to say that this is a relatively high value case, that it is important to the claimants or that it involves allegations of fraud (and, in relation to fraud, the court could not take a view on the merits of the allegations).

Following State of Qatar v Banque Havilland SA [2020] EWHC 1248, Model E disclosure needs to be linked to the specific issues and with some explanation provided as to the nature of the enquiry envisaged. It was not sufficient for the claimants to refer to, in effect, all communications by the defendants with any and all parties involved. That did not justify a “train of enquiry” order.

Just because the parties might be able to afford to carry out Model E disclosure did not mean it would be proportionate having regard to the court’s Overriding Objective (ie to deal with matters justly and at proportionate cost).

Control of server data

The claimants sought a declaration that data on the relevant server was in the “control” of the defendants for the purposes of disclosure. One of the defendants (Mr Braid) remained a director of the purchaser, which held the server. The other defendant (Mr Baker) was a former director and had certain contractual rights to be provided with access to documents, but it was said that these rights of access were “not however unlimited” and were “subject to exceptions”.

The defendants solicitors had written to the purchaser’s solicitors asking that the purchaser make available for the purposes of the defendants’ disclosure the data and documents held by the purchaser. The purchaser’s solicitors confirmed their willingness to do so subject to agreeing “an appropriate protocol”, and conditional on being indemnified in respect of costs by the claimant.

Moulder J noted that, although the court had been taken to various authorities on the question of “control” (including Pipia, referred to above, and North Shore Ventures Ltd v Anstead Holdings Inc [2012] EWCA Civ 11 (considered here)), it was clear (and indeed common ground) that the issue of “control” for these purposes was a question of fact. And since Pipia related to a parent/subsidiary relationship and North Shore Ventures related to beneficiaries under a trust, the judge considered them to be of limited assistance to the question of fact in this case.

The judge concluded that the claimants had not established that Mr Braid had “control” of the server data. He was one of six directors and was seeking to access the documents in his personal capacity. It was not a one-man company, and there was no evidence to justify the conclusion that Mr Braid could access the servers for the purposes of the litigation against him personally. The company was separately represented as a distinct legal entity and its lawyers had refused to provide access except on certain terms.

As regards Mr Baker, the evidence was that he retained certain rights of access to documents for the purpose of these proceedings, but the company was only obliged to use reasonable endeavours to make emails and group records available to Mr Baker and there was an express carve out for privileged documents. He did not have the right to obtain a copy of the entire servers. The judge said that, while Mr Baker had not provided copies of the relevant contractual arrangements referred to, she was not satisfied that the contractual rights amounted to control of the server data for the purposes of PD 51U, as the evidence did not support a finding of free and unfettered access and thus a finding of “control” as a matter of contract.