A recent disclosure ruling in Newman v Rightmove shows how familiar privilege rules can create unfamiliar difficulties in opt-out collective proceedings. The Competition Appeal Tribunal’s conclusion – that a proposed class representative (PCR) cannot rely on privileged communications while withholding the underlying material – applies familiar privilege waiver principles. Its implications for absent class members who are reluctant to speak openly may be less straightforward.
The PCR, Mr Newman, seeks to bring an opt-out claim on behalf of approximately 7,200 estate agents and other property businesses. The claim alleges that Rightmove abused a dominant position in the online property portal market. In support of the proposed opt-out structure, witness evidence referred to communications with proposed class members who were said to be concerned about adverse commercial consequences if they became publicly associated with the proceedings. That evidence was relied on to argue that requiring class members to opt in would deter participation. Rightmove denies any suggestion that it would retaliate against businesses supporting the claim.
Rightmove sought disclosure of the underlying communications. The Tribunal held that privilege had been waived: the PCR had referred to the communications and positively relied on them in support of an issue that the Tribunal will have to decide at certification. Fairness required Rightmove to be able to test the evidence rather than being presented with what it regarded as an opaque and potentially unrepresentative summary. As the Tribunal put it, the PCR could not “have it both ways”.
An orthodox answer—and an uncomfortable result
As a matter of conventional privilege doctrine, the result is readily understandable. A party cannot deploy favourable accounts of privileged communications while preventing scrutiny of the underlying material. The difficulty is that those principles were developed for ordinary litigation, not for collective claims involving absent class members whose confidentiality interests may not be identical to those of the PCR.
Rightmove’s concern was that where the prevalence and nature of class-member concerns are relied upon to support opt-out certification, it must have a fair opportunity to test that evidence. The difficulty lies not in recognising that entitlement, but in finding a procedure that preserves it without discouraging the kinds of engagement that can shed light on questions that are relevant to practicability in appropriate cases. The result is uncomfortable when viewed through the lens of collective redress policy.
The opt-out regime exists because individual consumers and businesses may lack both the economic incentive and the practical willingness to bring proceedings themselves. It is intended to improve redress and deter anti-competitive conduct without requiring class members to take proactive steps. That rationale may be particularly important where class members remain commercially dependent on the proposed defendant and fear damaging that relationship by becoming publicly associated with a claim.
Newman exposes a difficulty. Evidence of such concerns may support the case for an opt-out structure, but relying on particular class-member communications may require those communications to be disclosed so that the respondent can test their prevalence, context and weight. The Tribunal recognised the potential chilling effect, but held that it arose from the way the PCR had chosen to deploy the evidence. It limited disclosure to a lawyers-only confidentiality ring and allowed the PCR to withdraw the evidence instead, which the PCR subsequently did.
Those safeguards reduce the risk of wider disclosure, but may not eliminate the deterrent effect. A class member may still be reluctant to speak candidly if doing so could result in its identity and commercial circumstances being disclosed even within a confidentiality ring. The ruling should, however, enable claimant lawyers to explain that risk more clearly and to consider at an early stage whether the necessary evidence can be gathered in a form that does not entail reliance on identifiable privileged communications.
Of course each case turns on its own facts and there is no general requirement for the identities of class members who have spoken to the class representative to be provided to the defendant. For example, in a September 2025 disclosure ruling in Rodger v Google, the Tribunal refused Google’s request for the identities of four developers the class representative had consulted. It recognised that requiring participants to be identified could discourage engagement and inhibit open dialogue, and referred to the “sound public policy reasons” for withholding their identities absent “compelling reasons” for disclosure in the particular case (at [168]). The circumstances differed from Newman: the class representative had not relied on communications with those developers to establish a contested factual proposition. Nevertheless, the ruling is notable for its recognition of the value of class-member consultation, the practical difficulty of securing participation and the potential importance of preserving participants’ confidentiality.
Is there a middle course?
Anonymisation was rejected in Newman because identity was considered relevant to context: the same concern may carry different weight depending on the size and dependence of the business involved.
That does not mean the only option in future cases is full disclosure or not being able to make the point at all. But other approaches – such as surveys, sampling, or staged disclosure where identifying details are only revealed if necessary – all bring their own difficulties (particularly pre-CPO).
The policy tension therefore remains: if evidence of participation barriers cannot safely be gathered or used, those barriers risk being excluded from certification analysis – not because they do not exist, but because respondents are unwilling to be identified.
Defendant communications with class members
Two related decisions provide further context on the privileged treatment of communications with class members.
In McLaren, the Court of Appeal rejected the proposition that the Tribunal Rules contain a general prohibition on defendants communicating directly with class members. Communications with class members for the purpose of gathering evidence to defend the claim may attract litigation privilege where they are made for the dominant purpose of the litigation. Requiring defendants to disclose the details of those approaches to the class representative could reveal their evidence-gathering strategy or unused material. The Court therefore considered that such scrutiny could place defendants in the position of having to choose between preserving privilege and pursuing relevant evidence.
The principle was considered more recently in Ennis v Apple. Apple indicated that it intended to approach certain class members to seek voluntary disclosure, without copying the class representative or obtaining the Tribunal’s prior permission. The Tribunal held that its case-management powers could not be exercised in a manner incompatible with any litigation privilege attaching to those communications. Procedural directions governing communications with class members were the subject of a further hearing on 27 July 2026, in which the class representative argued that Apple had waived privilege by telling the Tribunal the questions it intended to ask class members. A ruling is awaited.
McLaren and Ennis therefore concern the protection of communications made while defendants are gathering evidence. Newmanaddresses a different stage of the analysis: the consequences where a class representative positively relies on privileged communications. Read together, the decisions illustrate that the treatment of class-member communications depends both on the purpose for which they were created and on how they are subsequently used in the proceedings.
Privilege at the boundaries of group litigation
While not an opt-out case, and so less relevant to untested questions on how privilege applies to passive class members, the recent Uber London Ltd v White ruling will be of interest to those working on group litigation.
In that case the Commercial Court considered communications generated while solicitors were instructed by a prospective funder to investigate potential claims on behalf of black cab drivers. The Court held that litigation privilege did not apply merely because the funder was assessing the merits of contemplated litigation. On the evidence, the dominant purpose was to enable the funder to decide whether to fund someone else’s claim—not to conduct its own litigation. Some documents could still attract legal advice privilege, but that required assessment document by document.
The distinction between Uber and Newman is important. In Newman, privilege was lost because of reliance on the material. In Uber, it did not arise in the first place.
Takeaway
The cases discussed above highlight a broader point: collective litigation depends on communications between class members, representatives, lawyers (claimant and defendant) and funders, but privilege does not automatically attach as information moves between them. And even if it does attach, it can be waived. Its protection depends on purpose, context, and subsequent use.
The immediate lessons are practical:
- Decide before filing whether class-member communications will be relied upon.
- Avoid summarising privileged communications to prove a disputed fact unless waiver has been considered.
- Consider surveys, aggregated evidence or other methods that do not depend on identifiable communications.
- Explain the confidentiality limits to prospective class members at the outset, including in relation to their identity specifically.
The harder question is whether ordinary waiver principles are fully equipped for a regime that depends on absent class members being willing to speak candidly in the first place.
* Geradin Partners acts for Professor Rodger in his claim against Google on behalf of UK app developers.
Leave a Reply