Key Non‑Party Actor Must Be Treated as a “Custodian” for Discovery, and Discovery Must Extend Beyond a Transaction Date Where the Category Is Open‑Ended

Case: National Paediatric Hospital Development Board v BAM Building Ltd (Approved) (Rev1) [2026] IEHC 465

Court: High Court of Ireland (Quinn J.)

Date: 14 July 2026

Topic: Further and better discovery; custodians; discovery from agents/advisers; eDiscovery/CAL; privilege; temporal scope of discovery.

1) Introduction

This judgment concerns an interlocutory dispute about compliance with a second discovery order in Commercial Court proceedings arising from the construction contract for the new children’s hospital at the St. James’s Hospital campus. The plaintiff employer seeks a declaration that an instruction dated 8 January 2019 (the “Phase B Instruction”) validly required the contractor to commence Phase B works. The defendant contractor disputes validity, including (after amendment) on the basis that the individual who issued the Phase B Instruction, Clare White (a director of O’Connell Mahon Architects (“OCMA”)), was not appointed as the contract’s Employer’s Representative at the time.

The Second Discovery Order (Mulcahy J., 29 October 2025) required additional discovery focused on the appointment/substitution of the Employer’s Representative, and the BDP–OCMA sub-consultancy arrangements for Employer’s Representative services. The defendant alleged that the plaintiff’s subsequent discovery was deficient in (i) custodian selection (notably, not treating Ms White as a custodian), (ii) the small volume of disclosed documents, (iii) reliance on documents sourced from BDP/OCMA rather than the plaintiff’s own records, and (iv) failure to re-review previously discovered material and/or to cover the correct date ranges—particularly for categories without express date limits.

2) Summary of the Judgment

Quinn J. applied the established test for further and better discovery and made two targeted orders:

  1. Ms White must be treated as a custodian: The plaintiff must identify Ms White as a custodian and collect, review and discover (by a sixth affidavit) all documents held by her personally and/or in her capacity as Employer’s Representative and/or as a director of OCMA, falling within the Second Discovery Categories.
  2. Post–8 August 2023 documents must be discovered: The plaintiff must review, collect and discover all documents coming into existence on or after 8 August 2023 that fall within the Second Discovery Categories.

All other reliefs were refused. In particular, the Court declined to order broader “process” explanations (beyond what was already sworn), declined to infer inadequacy from low document counts, and declined to require re-discovery or cross-referencing of already-discovered documents.

3) Analysis

3.1 Precedents Cited

(a) Sterling Winthrop Group Limited v. Farbenfabriken Bayer [1967] IR 97

Quinn J. treated Kenny J.’s decision as the “seminal” authority and set out the familiar principles governing further and better discovery. The key points emphasised were:

  • The Court has power to order a further affidavit of discovery.
  • A bare assertion that more documents exist is insufficient.
  • A further affidavit may be ordered where, from pleadings/affidavits/discovered documents/admissions, it appears relevant documents exist but have not been disclosed.
  • The Court may act where discovered material gives grounds to suspect further relevant documents exist, or where there are reasonable grounds to believe the deponent misunderstood the issues (leading to omission).
  • The Court should not order further discovery unless it is shown that relevant documents exist and are undisclosed, or that the deponent’s relevance assessment is wrong / issues misunderstood.

How it influenced the decision: The Court used Sterling Winthrop as the benchmark to reject most of the defendant’s complaints (e.g., “gaps” and low volume), while still making discrete, issue-driven orders where the Court identified concrete inadequacies or unexplained temporal cut-offs.

(b) Daly v. Ardstone Capital Limited [2020] IEHC 200 and Victoria Hall Management Ltd & Ors. v. Cox & Ors. [2019] IEHC 639; Kelland Homes Limited v Ballytherm Limited & Ors. [2019] IEHC 46

Quinn J. relied on Murray J.’s articulation in Daly v. Ardstone Capital Limited that the Sterling Winthrop factors are not exhaustive and that the Court may consider additional evidence (from persons able to properly aver to it) to evaluate whether discovery was properly undertaken. He also discussed Murray J.’s recognition that the Court may order a more limited affidavit addressing specific concerns rather than ordering further discovery “per se”, with Victoria Hall and Kelland Homes as examples.

Notable clarification: Quinn J. rejected the plaintiff’s submission that this “limited affidavit” jurisdiction only arises after the Sterling Winthrop test is satisfied. He interpreted Ardstone as permitting targeted explanatory orders where genuine concerns exist about the adequacy of the discovery process, even if non-compliance is not conclusively established.

(c) Linfen Limited v. Rocca [2009] IEHC 292

Quinn J. referenced MacMenamin J.’s distinction between (i) documents generated by an agent that “belong to” the principal (and so may be within the principal’s discoverable control), and (ii) documents created by professional advisers for their own assistance. This was relevant to the defendant’s concern about BDP/OCMA documents being provided on a “confidential” basis and the “client file” versus non-client file material.

How it influenced the decision: The Court examined whether anything in the affidavits suggested that documents had been withheld from the plaintiff’s solicitors due to their character (agent/adviser material or privilege). Quinn J. found no evidential basis to infer that relevant documents were excluded from review merely because they were described as outside the “client file”.

3.2 Legal Reasoning

(a) Custodian selection: general deference, but not where the individual is central

The Court accepted that, as a general principle, a party is not obliged to provide an analysis justifying why certain individuals were not selected as custodians, and the defendant’s broad assertions about other individuals being “directly involved” were insufficient to displace the plaintiff’s sworn evidence on its custodian approach.

However, Ms White’s position was materially different: she authored the Phase B Instruction and was expressly named across all four categories in the Second Discovery Order. The plaintiff had sought and received documents from OCMA as an “agent/adviser”, but Quinn J. drew an important procedural distinction:

  • Firm-wide agent/adviser collection (requesting documents from OCMA/BDP) may be diligent, but it is not necessarily as rigorous as…
  • Custodian-based collection (interrogating an individual’s emails/devices/hard copy records).

Because Ms White was central and there was no impediment to treating her as a custodian (notwithstanding she was not an employee of the plaintiff), her exclusion from the custodian process was “not warranted” in the circumstances. This justified an order specifically directing her inclusion as a custodian and a corresponding further affidavit.

(b) “Gaps” and low document counts: discovery is not an explanation engine

The defendant pointed to the small number of documents discovered (91 produced; 11 withheld as privileged) over a long period and argued that the result was “implausible”, particularly given the use of continuous active learning (CAL) on a large review set.

Quinn J. refused to equate “surprising scarcity” with non-compliance. Critically, he held that the defendant’s complaint—essentially asking what “prompted or triggered” the belated execution of the OCMA sub-consultancy agreement—misunderstood the purpose of discovery. Discovery is for producing relevant documents that exist and are within the party’s discovery obligations; it is not designed to answer “why” questions, nor to guarantee that all narrative gaps are filled.

On the affidavits before the Court, the discovery process (including CAL, validation measures, and solicitor oversight) was described with sufficient care that the Sterling Winthrop threshold for a further order on this point was not met.

(c) Documents sourced from BDP/OCMA: no adverse inference without evidential foundation

The Court accepted that, given the subject matter of the second categories (appointment/substitution and sub-consultancy arrangements for Employer’s Representative services), it was unsurprising that many responsive documents would originate from BDP/OCMA. The fact that most documents came from those entities did not, without more, demonstrate that the plaintiff’s own records were incomplete or that searches were inadequate.

(d) Interaction between first and second discovery; date ranges; and the post–8 August 2023 issue

The defendant sought orders requiring (i) review and discovery of all documents responsive to the Second Discovery Categories even if previously discovered, or (ii) at least cross-referencing previously discovered documents. Quinn J. declined these orders, construing the affidavits as showing that the plaintiff’s second discovery was not improperly constrained by what had been “previously collected”.

But Quinn J. identified a discrete temporal deficiency: the latest discovered document was 8 August 2023 (the date of signing of the OCMA sub-consultancy agreement), yet the defendant’s amendment application (and the controversy about authority/appointment) arose later. The Court considered it “extraordinary” if there were no post–8 August 2023 communications (internal or external), some of which might be non-privileged and some privileged (and thus discoverable in the privilege schedule). No explanation was offered for the absence of such material.

This justified an order directing review and discovery of all documents from 8 August 2023 onward matching the four Second Discovery Categories.

3.3 Impact

  • Custodian discipline extends to key non-party actors: Where an individual outside the litigant’s organisation is central to the pleaded issues (e.g., the author of a disputed instruction), “agent/adviser” production may not be an adequate substitute for custodian-level collection. This is a practical, litigation-management precedent in large projects where functions are outsourced.
  • Courts will resist “discovery by incredulity”: Low yield, even over long time spans and even where CAL is used, will not automatically satisfy Sterling Winthrop. Applicants must connect suspicions to evidence in a way recognised by the authorities.
  • Open-ended categories can require post-event discovery: Even where a dispute focuses on an historic act (here, January 2019), if the discovery categories are framed without a cut-off (as Categories 3 and 4 were), the producing party may be required to search forward in time—particularly where later communications likely exist about the very issue that triggered amended pleadings and additional discovery.
  • Affidavits addressing concerns: The judgment supports a flexible reading of Daly v. Ardstone Capital Limited that permits targeted affidavits/explanations where appropriate, without treating such directions as strictly contingent on first proving full Sterling Winthrop non-compliance.

4) Complex Concepts Simplified

  • Discovery: A process requiring parties to list (and often produce) documents relevant to pleaded issues, subject to privilege and other recognised limits.
  • Further and better discovery: A follow-up order made where the Court is satisfied that the first discovery has likely omitted relevant material or was improperly approached.
  • Custodian: A person whose email accounts, devices, and records are searched as a primary source of potentially relevant documents (a standard approach in eDiscovery).
  • Agent/adviser collection vs custodian collection: Requesting documents from a firm (e.g., architects/consultants) is different from conducting a custodian-level search of a named individual’s accounts/devices; the latter is typically more granular and systematic.
  • CAL (continuous active learning): Technology-assisted review that prioritises likely-relevant documents for human review. The Court treated CAL as part of the described process but did not accept that a low yield alone proves malfunction or non-compliance.
  • Privilege: Certain communications (commonly lawyer–client and litigation-related materials) may be withheld from production but must generally be listed with the nature of the privilege claimed.

5) Conclusion

National Paediatric Hospital Development Board v BAM Building Ltd [2026] IEHC 465 is a significant procedural decision on targeted further discovery in complex, consultant-heavy construction disputes. Quinn J. reaffirmed the discipline of Sterling Winthrop Group Limited v. Farbenfabriken Bayer while making two practical, high-impact orders: (i) a central non-employee actor (the author of the disputed instruction) must be treated as a custodian for discovery purposes, and (ii) where discovery categories are not time-limited, discovery may need to extend beyond the date of a key agreement to capture later communications that likely exist and may be relevant (or privileged and thus listable). The judgment also underscores that discovery is not a mechanism to force narrative explanations for perceived “gaps”, but a structured obligation to identify and disclose relevant documents within the proper legal parameters.