Amending to Plead Breach of Confidence Requires a Coherent, Particularised Case: Titan Wealth Services Ltd & Anor v Tavistock Investments PLC & Ors [2026] EWCA Civ 500

1. Introduction

This Court of Appeal decision concerns a case-management question with significant practical consequences for breach of confidence and trade secrets litigation: when (and on what standard) a party may amend pleadings to introduce a counterclaim for misuse of confidential information.

The appellants were Titan Wealth Services Limited (a Jersey company providing wealth management and administration services) and Titan Asset Management Limited (“Titan AM”, a UK investment manager). Titan AM was formerly part of the Tavistock group (then “Tavistock Wealth”) and managed the “Acumen Funds”.

The respondents included Tavistock Investments PLC (the holding company), Tavistock Asset Management Ltd (“Tavistock AM”), and certain directors. Under a 2021 Share Purchase Agreement (“SPA”), Titan acquired Tavistock Wealth (later Titan AM). Tavistock retained its own Model Portfolio Service (“MPS”) and entered an Outsourced Management Agreement (“OMA”) with Titan AM (September 2021), under which Titan AM provided investment management services for Tavistock AM’s MPS. Clause 5 of the OMA imposed confidentiality obligations in respect of each party’s “Confidential Information”, defined as information disclosed by one party to the other in connection with the OMA.

Against the background of a wider commercial dispute about the decline of the Acumen Funds and alleged breaches of the SPA, Tavistock sought to amend its Defence and Counterclaim to add (i) copyright infringement and (ii) breach of confidence / trade secrets claims, alleging that Titan AM launched a competing Titan MPS using Tavistock’s confidential information. The High Court (Bryan J) allowed the amendments. Titan appealed only the permission to amend to add the breach of confidence counterclaim.

2. Summary of the Judgment

The Court of Appeal (Males LJ, with Miles LJ and Asplin LJ agreeing) allowed the appeal and held that the judge should not have permitted the amendment introducing the breach of confidence counterclaim.

The proposed pleading was found to be:

  • incoherent and self-contradictory (notably, it alleged Tavistock communicated certain information to Titan AM, yet the pleaded “examples” were emails sent by Titan AM to third parties);
  • insufficiently particularised as to what information was said to be confidential (“Everything that is not published about the Tavistock MPS is confidential”); and
  • unsupported by the very materials relied upon to justify an inference of misuse (the examples of factsheets/holdings did not support the pleaded inference).

The Court emphasised that it was not deciding that Tavistock could never plead a viable claim; rather, Tavistock could apply again with a more focused, properly particularised case, potentially after further analysis of factsheets due for disclosure in July 2026. The Court declined to consider a revised amendment proposed after circulation of the draft appellate judgment.

3. Analysis

3.1 Precedents Cited

(a) The amendment/summary judgment threshold and limits on merits review

The Court reiterated that a “necessary condition” for an amendment is that the amended claim has a real prospect of success, aligned with the summary judgment test. Relying on CNM Estates (Tolworth Tower) Ltd v Carvill-Biggs [2023] EWCA Civ 480, [2023] 1 WLR 4335, the Court accepted the general proposition that—particularly for amendments which are not late/very late—the court should not ordinarily conduct a mini-trial of the claim’s strength.

However, the Court relied on Elite Property Holdings Ltd v Barclays Bank Plc [2019] EWCA Civ 204 and Kawasaki Kisen Kaisha Ltd v James Kemball Ltd [2021] EWCA Civ 33, [2021] 3 All ER 978 to confirm an important qualification: the court may refuse an amendment where the pleaded facts are implausible, self-contradictory, not supported by contemporaneous documents, or where the proposed pleading is not coherent and properly particularised.

(b) Particularisation in breach of confidence / trade secret claims

The Court’s central doctrinal anchor on pleading standards was Celgard LLC v Shenzhen Senior Technology Material Co Ltd [2020] EWCA Civ 1293, [2021] FSR 1, in which Arnold LJ approved Laddie J’s well-known insistence on specificity in confidentiality claims.

In turn, the Court quoted extensively from Ocular Sciences Ltd v Aspect Vision Care Ltd [1997] RPC 289, which explained why unparticularised confidence claims can be oppressive and why defendants must know what case they have to meet (including for public domain challenges and for workable injunctions). The Court also referred to John Zink Co. Ltd. v. Wilkinson [1973] RPC 717 and John Zink Co. Ltd. v. Lloyds Bank Ltd [1975] RPC 385 as authorities illustrating the court’s power to require particulars and to strike out where litigation becomes abusive.

Finally, CMI-Centers for Medical Innovation GmbH v Phytopharm plc [1999] FSR 235 was cited to emphasise that even at interlocutory stages the defendant must be told “precisely” what information is said to be confidential and misused.

3.2 Legal Reasoning

(a) Focus on the pleading itself as a coherent, self-contained case

The Court stressed that the assessment must be anchored in the proposed pleading, which must stand as a “coherent and self-contained statement” of the case. The Court declined to “reformulate” or “salvage” parts of a broadly drafted claim where the pleading, as advanced, failed basic coherence and clarity tests.

(b) A contract-defined confidentiality obligation cannot be assumed to cover “everything”

A key interpretive point was the OMA definition of “Confidential Information”: information “disclosed by either Party ... to the other Party ... in connection with this agreement”. Males LJ considered it “reasonably clear” that Tavistock’s contractual ‘Confidential Information’ is limited to information disclosed by Tavistock to Titan.

This mattered because Tavistock’s pleaded theory depended on the premise that “Tavistock AM communicated information necessary to implement the Tavistock MPS to Titan AM”. Yet the pleaded “examples” (the “Investment Emails”) were not communications from Tavistock to Titan AM at all; they were communications from Titan AM to third parties arranging onboarding. The pleading therefore collapsed under internal contradiction: it relied on “Tavistock-to-Titan” disclosure while exhibiting “Titan-to-third-party” documents.

(c) Over-breadth and non-definition: “Everything not published is confidential”

Tavistock attempted to define “Tavistock MPS Information” as including (i) “Investment Information”, (ii) the Model Portfolios, and (iii) “each part of the Tavistock MPS”, “save insofar as ... deliberately published”. The Court held that this definition was broad and undefined, leaving Titan unable to know what information was actually in issue.

The Court accepted that some information might be properly confidential (for example, the full composition of model portfolios beyond the limited disclosures in publicly available factsheets), but held that Tavistock’s deliberate decision to frame its claim at maximal breadth rendered it unsuitable for permission.

(d) The pleaded “misuse” inference was not supported by the examples relied on

Tavistock pleaded an inference of misuse from:

  • onboarding of the same investments for Titan as for Tavistock (but this was done openly and emails were copied to Tavistock);
  • similarities between Titan and Tavistock factsheets (but factsheets were public, so similarities did not support misuse of confidential information, even if they supported copyright infringement);
  • similarities in underlying investments (potentially powerful if established over time), but Tavistock’s three examples did not support the inference; one example actively contradicted it.

The Court signalled what a viable misuse case might look like (e.g., mirrored portfolio changes over time; persistent overlap in top holdings; comparable asset allocations), but held that the case as pleaded did not reach a coherent “real prospect” threshold.

(e) Distinguishing a potentially arguable unpleaded theory from the pleaded one

Tavistock argued (in submissions) that even if confidential information was generated by Titan AM while performing the OMA, Titan AM owed duties of confidence/good faith preventing it from using that work product outside the OMA. The Court acknowledged that a claim along these lines “may” be pleadable with real prospects, but held that this was not the case actually pleaded, which remained anchored to the (incorrect) proposition of disclosure from Tavistock to Titan AM.

(f) Particularisation and proportionality: guarding against oppressive disclosure

Echoing the policy in Ocular Sciences Ltd v Aspect Vision Care Ltd, the Court underlined that breach of confidence claims can generate wide-ranging and expensive disclosure, risking disproportionate cost relative to the issues at stake—particularly where the allegedly competing product operated only for a short period. The Court rejected the attempt to reverse the burden (“it would be simple for Titan to prove it didn’t copy”) and insisted that Tavistock must first plead a viable, particularised case.

3.3 Impact

  • Sharper pleading discipline for confidence/trade secret amendments: The decision reinforces that a party cannot obtain permission to amend on the basis of a deliberately broad formula (“everything not published is confidential”) and then seek to discover whether any misuse occurred. Amendments must articulate a coherent theory of confidentiality, identify the relevant information with reasonable specificity, and avoid internal contradictions.
  • Contractual definitions matter: Where an NDA/confidentiality clause defines “Confidential Information” by reference to “information disclosed” by one party to another, pleadings must align with the actual direction and mechanism of disclosure. Attempts to treat the contract as capturing all information “generated” during performance may require a distinctly pleaded contractual construction, an implied term argument, or a separate equitable duty theory.
  • Separating public-document copying from confidential misuse: The judgment draws a practical evidential boundary: copying public factsheets may support a copyright claim, but it does not itself ground an inference of misuse of confidential information.
  • Guidance for outsourced investment/financial services arrangements: In outsourced management contexts, the “confidential information” question often concerns whether outputs (research, recommendations, portfolio construction) are owned/confidential to the client or remain part of the provider’s skill and know-how. This decision signals that such disputes must be pleaded with precision and rooted in the contract’s actual structure and information flows.

4. Complex Concepts Simplified

  • “Real prospect of success”: The threshold used for both summary judgment and (as a necessary condition here) permission to amend. A claim must be more than arguable in the abstract; it must be viable on its pleaded facts.
  • “Pleading” and “particularisation”: The pleading is the formal statement of the case. Particularisation means identifying, with sufficient detail, what information is said to be confidential, how it was communicated/obtained, and what misuse is alleged—so the defendant can respond and the court can police scope and proportionality.
  • Breach of confidence vs trade secrets: “Breach of confidence” is a common law/equitable doctrine protecting confidential information. “Trade secrets” (here pleaded under the Trade Secrets Regulations) require, broadly, that the information is secret, valuable because secret, and subject to reasonable steps to keep it secret. Both require careful identification of the information.
  • “Onboarding” (in this case): An administrative process to make an investment product available for purchase; it does not itself imply a decision to invest or reveal the value of an investment strategy.
  • Copyright vs confidentiality: Copyright protects original expression (text/design/layout), even if publicly available. Confidentiality protects secrecy. Copying a public document can infringe copyright without involving any misuse of secrets.

5. Conclusion

Titan Wealth v Tavistock [2026] EWCA Civ 500 is a focused but important appellate intervention in amendment practice for breach of confidence and trade secrets claims. It confirms that, notwithstanding the court’s general reluctance to evaluate merits on amendments, permission should be refused where the proposed claim is internally incoherent, self-contradictory, and inadequately particularised.

The judgment’s broader significance lies in its insistence that confidentiality litigation must begin with disciplined identification of the alleged secrets and a coherent account of how they were disclosed/obtained and misused—particularly where the alternative is expansive, costly disclosure driven by an ill-defined complaint.