Open justice first: High Court sets exacting, evidence-specific test for permanent confidentiality orders and judgment redactions in patent litigation

Introduction

In Sandoz AG & Rowex Ltd v Bayer Intellectual Property GmbH [No. 2] [2025] IEHC 500, the High Court (Mulcahy J) refused a suite of post-trial confidentiality measures sought by Bayer following a successful challenge by Sandoz and Rowex to the validity of Irish patent EP (IE) 1 845 961 relating to the once-daily dosing regimen for rivaroxaban (marketed as Xarelto).

The application arose after the Court had delivered its substantive judgment to the parties on 25 June 2025 but before wider publication. At trial, Bayer had provided discovery under a confidentiality agreement and obtained an interim order preserving confidentiality pending a final determination. Post-judgment, Bayer sought to (i) redact specified passages in the judgment (five redactions), (ii) redact extensive portions of the public transcript referring to 19 documents, and (iii) deem 25 of 31 “opened” confidential discovery documents to retain confidentiality notwithstanding their deployment at trial.

The core question was: when, and on what evidential basis, may an Irish court depart from the constitutional requirement that justice be administered in public to protect commercial confidentiality, particularly in patent litigation? The Court’s answer places the onus firmly on the party seeking protection to adduce specific, document-linked evidence of a clear interest and pressing circumstances showing a real risk of harm—general assertions will not suffice.

Summary of the Judgment

Mulcahy J refused all elements of Bayer’s application:

• Redactions to the judgment were refused. The Court found no cogent evidential basis to show that the proposed redacted passages disclosed confidential information beyond what was already otherwise disclosed or publicly known.
• Redactions to the transcript were refused. The Court emphasised that Bayer had advanced only generic assertions and had not made a specific, document-by-document showing of risk of harm.
• Orders deeming trial-opened documents to retain confidentiality were refused. The Court would not conduct its own granular review in the absence of the necessary particularised evidence from Bayer identifying protectable material and explaining the harm from disclosure.

Applying the Supreme Court’s framework in Gilchrist and Rogers v Sunday Newspapers, the Court held that, absent a statutory basis, departures from open justice require “very clear” interests and “pressing” circumstances; the Court must approach such claims with “resolute scepticism” and adopt the minimum departure necessary. Bayer failed to establish a real risk of harm or that the information qualified as a trade secret in any relevant sense. Notably, the Court found that the documents, in aggregate, reflected standard development processes that were expected by experts and had been publicly discussed by Bayer elsewhere, undermining claims of inherent confidentiality.

Procedurally, to accommodate a potential appeal, the judge proposed to publish the principal judgment with Bayer’s proposed redactions pending any appellate determination, and to stay the discharge of the interim confidentiality order over the transcripts and documents for the same period. On costs, the Court indicated that Sandoz, having successfully resisted the application, is entitled to its costs of the application.

Analysis

Precedents Cited and Their Influence

The Court anchored its analysis in Irish constitutional principle and leading authority, with measured reference to English case law as persuasive but not controlling.

• Article 34.1 of the Constitution: Establishes the baseline—justice must be administered in public, with statutory exceptions permitted “in such special and limited cases as may be prescribed by law.”
• Section 45(1)(d) of the Courts (Supplemental Provisions) Act 1961: Provides a specific statutory exception for proceedings involving the disclosure of a “secret manufacturing process.” Bayer did not contend that its documents disclosed a secret manufacturing process; accordingly, s.45(1)(d) could not be invoked.
• Gilchrist and Rogers v Sunday Newspapers [2017] IESC 18; [2017] 2 IR 284: The seminal authority on the inherent jurisdiction to depart from open justice in the absence of a statutory basis. O’Donnell J set out controlling principles, including:
  • Open justice is a fundamental constitutional value;
  • Exceptions are strictly construed and must be no more than necessary;
  • Where relying on common law rather than statute, the interests must be very clear and the circumstances pressing;
  • The Court must adopt a posture of “resolute scepticism.”
  Mulcahy J treats Gilchrist as the governing framework. He expressly transposes its requirement for “very clear” interests and “pressing” circumstances to the commercial confidentiality context (paras 80, 83).
• Re R Ltd [1989] IR 126 and Irish Times v Ireland [1998] 1 IR 359: Historic markers on the limits of non-public trials. Gilchrist acknowledges that Irish Times softened the rigidity attributed to Re R, signalling that some restrictions can exist under inherent jurisdiction, but still subject to strict scrutiny.
• BOC Aviation Ltd v Lloyds [2024] IEHC 162: High Court applied Gilchrist in a different setting (witness safety). McDonald J first asked whether a real risk of harm was established; only then did he consider whether narrower measures sufficed. Mulcahy J adopts the same analytical sequence, substituting “commercial harm to confidential information” as the relevant interest (paras 89–90).
• Teva v Boehringer Ingelheim (HC, 2014 transcript): The High Court used s.45(1)(d) to hear certain evidence in camera where a secret manufacturing process was involved. This underlined that when the statutory exception applies, the route to privacy is clearer; otherwise, the Gilchrist inherent jurisdiction test governs.

Comparative English authorities were used as instructive on factors and techniques, but not as a legal foundation:

• CPR r.31.22(2) (England & Wales) permits orders preserving confidentiality in documents even if read in open court. Ireland has no equivalent procedural rule. The Court emphasised that, while instructive, there is “no direct read across” because Ireland’s constitutional architecture gives open justice express and entrenched force (para 82).
• Lilly Icos Ltd v Pfizer Ltd [2002] 1 WLR 2253: Stresses the centrality of open justice, that general assertions of harm are insufficient, and that the role/centrality of the document to the trial matters. Mulcahy J drew on this for the need for “very good reasons,” specificity, and the weight to be given to core trial documents.
• Smith & Nephew plc v Convatec [2014] EWHC 146 (Pat): Birss J granted protection for genuinely technical trade secrets necessary for adjudication (manufacturing process), but refused protection for regulatory material not linked to protectable know-how. Supports a targeted, category-specific approach requiring evidence tied to the documents.
• JC Bamford Excavators Ltd v Manitou [2024] Ch 215: Court of Appeal emphasised correctly characterising the information (e.g., “technical trade secrets”) and assessed the degree of confidentiality. Mulcahy J adopts the threshold exercise of characterising the information at issue and its confidentiality in fact (para 91).
• Unwired Planet v Huawei [2018] RPC 8: Redactions to judgments demand “powerful reasons, supported by cogent evidence.” Mulcahy J applies this standard to the proposed judgment redactions (para 101).
• Scott v Scott [1913] AC 417 and Home Office v Harman [1983] 1 AC 280: Classic statements of open justice cited in Lilly Icos (and referenced in Gilchrist), reinforcing that publicity is a cornerstone of the administration of justice.

Legal Reasoning

The Court’s reasoning proceeds in structured stages.

1. Jurisdictional foundation (paras 37–55):
   • Open justice is constitutional (Art. 34.1). There is no statutory hook, such as s.45(1)(d) (secret manufacturing process), that applies here.
   • Therefore, Bayer must rely on the inherent jurisdiction per Gilchrist, which allows exceptions only where “very clear” interests and “pressing” circumstances are shown, and only to the minimum extent necessary.
2. Characterise the information and identify the risk (paras 89–97):
   • Bayer’s affidavit evidence (Haselhorst) was generic—asserting that the documents reveal internal development processes and are “inherently confidential” and valuable to competitors.
   • The Court required specificity: identification of what in which document would cause what harm if published. Bayer did not provide document-by-document showings, despite seeking orders over 25 documents and wide transcript redactions.
   • Context undermined the confidentiality claim: many documents were first put into evidence by Bayer; some similar process information has been publicly discussed by Bayer in journals; and experts testified the processes reflected standard, expected development steps (para 92, 94).
   • Given the patent was found invalid for obviousness, the Court noted this does not negate confidentiality per se, but it undermined any inference that the development documents inherently embody inventive or secret know-how (para 85).
3. Trade Secrets Directive (paras 70–72, 95):
   • The Directive and Irish Regulations concern proceedings about unlawful disclosure/use of trade secrets. This case was not such a proceeding.
   • Even if relevant in principle, Bayer’s information did not satisfy Article 2(1)(a) (not generally known among relevant circles). The trial evidence showed that the processes disclosed were known or expected by those in the field and had been discussed in publications.
4. Balancing against open justice and the scope of any departure (paras 96–102):
   • Because Bayer failed to establish a real, pressing risk of harm, there was no basis to override open justice.
   • In any event, redactions to a judgment demand powerful reasons and cogent, document-specific evidence. Those were absent. The Court rejected the notion that selective redactions of “recitations” of documents could be justified when similar content appeared elsewhere unredacted (para 98).
   • The Court stressed it is not its role to conduct a roving, document-by-document confidentiality review where the moving party has not made the necessary specific case (para 99).
   • Argument that the documents were of “limited significance” failed; they were in fact central enough to be used extensively and cited in the judgment, mandating a very demanding approach to any confidentiality claim (paras 100–101).
5. Interim management and appellate comity (para 103):
   • To respect potential appellate intervention, the Court proposed to publish the judgment with Bayer’s proposed redactions pending appeal and to stay the discharge of the interim confidentiality order for transcripts and documents to the same point.

Impact

This judgment meaningfully clarifies the Irish approach to post-trial confidentiality in commercial and IP cases, particularly where discovery documents have been opened at trial:

• No CPR 31.22 analogue—exceptionality and evidence-based scrutiny in Ireland. Absent a rule-based jurisdiction akin to England and Wales, parties must meet the Gilchrist threshold under inherent jurisdiction. The constitutional imperative of open justice is paramount; departures are exceptional.
• Specific, document-linked evidence is mandatory. Generic affidavits asserting commercial sensitivity or confidentiality of “internal processes” will not justify redactions or post-trial confidentiality orders. Parties must identify particular passages, explain why they are confidential, and demonstrate the concrete harm likely to flow from publication.
• Judgment redactions face the highest bar. Redacting a reasoned judgment risks impairing public understanding of judicial reasoning. Courts will require “powerful reasons” backed by “cogent evidence.” Routine redaction requests in IP cases should not be expected to succeed.
• Open justice is not displaced by confidentiality agreements or interim orders. Such agreements are relevant background but cannot bind the Court or displace constitutional obligations. Interim protective orders are a pragmatic case-management tool, not a predictor of permanent relief.
• Trade secret claims must be substantiated and contextualised. Information long discussed in scientific publications or widely known among those in the field will not qualify as a trade secret. Parties should consider whether s.45(1)(d) (secret manufacturing process) truly applies; if not, they face a steep inherent-jurisdiction hill.
• Discovery “chilling effect” arguments carry little weight without evidence. The Court rejected speculative claims that refusing orders would deter voluntary discovery, noting that procedural fairness and open justice cannot be compromised on conjecture.
• Practical guidance for future litigants. If redactions or confidentiality orders are sought:
  • Lead evidence from knowledgeable personnel tying specific documents or passages to specific harms;
  • Explain the present-day value and secrecy measures, and why publication would cause material damage;
  • Propose narrowly tailored measures, demonstrating why lesser alternatives (e.g., anonymisation) would not suffice;
  • Acknowledge what is already public and distinguish the genuinely secret residue.

Complex Concepts Simplified

• Open justice: The constitutional rule that court hearings and reasons are public. It promotes legitimacy, accountability, and public understanding.
• In camera: A hearing conducted in private, excluding the public. Permissible only in special cases prescribed by law or under the Court’s limited inherent jurisdiction.
• Inherent jurisdiction (post-Gilchrist): The Court’s power to regulate its process to protect very clear interests in pressing circumstances, applying strict scrutiny and adopting the least restrictive measure.
• Confidentiality agreement vs. open justice: Parties can agree to treat documents as confidential inter se, but such agreements cannot override the public character of court proceedings or bind the Court’s constitutional duties.
• “Opened” documents: Documents introduced into evidence (e.g., via expert reports adopted in evidence or used in cross-examination) and thus, absent court order, typically treated as part of the public trial record.
• Trade secret (Directive 2016/943): Information that is secret, has commercial value because it is secret, and has been subject to reasonable secrecy measures. If the information is generally known in the relevant professional circles, it is not a trade secret.
• Obviousness/patent invalidity: A patent is invalid if the claimed invention would have been obvious to a person skilled in the art at the priority date. A finding of obviousness does not automatically negate confidentiality of development documents, but it weakens any inference that such documents necessarily embody valuable, secret inventive know-how.

Conclusion

Sandoz v Bayer [No. 2] is now a leading Irish authority confirming that permanent confidentiality orders and judgment redactions in IP litigation will be granted, if at all, only on a strictly exceptional, evidence-specific basis under the Court’s inherent jurisdiction. The decision cements five practical propositions:

1. Open justice is paramount and departures are exceptional.
2. Generic assertions of commercial sensitivity are insufficient; document-specific evidence and a clear showing of pressing, real risk of harm are required.
3. Trade secret labelling must be borne out by current realities—if the professional community already knows the substance, protection is unlikely.
4. Judgment redactions demand the strongest justification and cogent evidence.
5. Courts will not conduct a roving confidentiality assessment; the moving party must do the heavy lifting in its application.

The ruling aligns Irish practice with the spirit—though not the rule-based mechanics—of English case law by endorsing interim protective orders as a pragmatic tool, while insisting that permanent constraints on publicity require compelling, narrowly tailored justification. For patent and trade secret litigants, the message is clear: prepare highly particularised evidence tied to specific materials and harms, or expect open justice to prevail.