A partial green light: What the Apple/Amazon judgment means for future collective actions

On 28 September 2026, the UK Competition Appeal Tribunal (CAT), in JLP A&A Class Representative Limited v Apple Inc. and Others [2026] CAT 75, certified part of a proposed collective action against Apple and Amazon concerning alleged anti-competitive agreements relating to the sale of Apple products through Amazon's marketplace. 

The CAT granted conditional certification to claims relating to purchases of Apple products through Amazon's marketplace (the On-Amazon Claim), while refusing to certify a broader claim concerning purchases through other retail channels (the Off-Amazon Claim).

The judgment is notable not only for the CAT’s partial certification of the claim, but also because it provides important guidance on expert independence, distribution mechanisms, funding arrangements and class representative remuneration: 

• By certifying only part of the claim, the CAT has confirmed that certification is not an all-or-nothing exercise. Viable claims may proceed even where other alleged theories of harm fall away. Here, despite the lower quantum, the CAT found that the On-Amazon Claim (with estimated damages of £289-306 million) passed the cost-benefit analysis, notwithstanding its refusal to certify the Off-Amazon Claim.
• Experts who have played a role in originating proposed claims can expect heightened scrutiny regarding their independence. 
• Distribution matters - practical and low-friction distribution proposals can materially strengthen an application for certification, and feed into the CAT’s cost-benefit analysis. 
• The CAT signalled its willingness to scrutinise (and if necessary, reject) funding arrangements at the certification stage in circumstances where such arrangements would limit any potential recovery by class members.
• The CAT continues to take an increasingly active role in supervising class representative remuneration, with £150 per hour and £750 per day emerging as important reference points.

Background

The judgment follows an earlier certification application brought by Professor Christine Riefa. In 2025, the CAT refused to certify that claim, finding that the proposed class representative, Professor Riefa, had not demonstrated sufficient independence and robustness to act on behalf of the class (see our earlier briefing here). 

Following that decision, Justin Le Patourel replaced Professor Riefa as the proposed class representative (formally as sole director of the special purpose vehicle acting as proposed class representative), and a fresh application was issued advancing largely the same allegations, with an updated claim period, as well as the addition of a claim for injunctive relief. 

The claim alleges that agreements entered into between Apple and Amazon in 2018 unlawfully restricted competition by limiting the sale and advertising of Apple products on Amazon’s marketplace, causing consumers to pay higher prices for Apple products purchased both on Amazon and through other retail channels. 

The proposed application advanced two distinct theories of harm:

• an On-Amazon Claim, alleging that the agreements reduced competition among sellers on Amazon's marketplace, leading to higher prices for purchases made through Amazon; and
• an Off-Amazon Claim, alleging that the same conduct increased prices across wider retail channels, including Apple's own stores and competing online and offline retailers.

In relation to the On-Amazon Claim, the CAT concluded that the class representative had advanced a credible theory of harm supported by a sufficiently developed methodology and preliminary economic evidence. Crucially, neither Apple nor Amazon advanced evidence challenging the proposed methodology, and neither Proposed Defendant identified fundamental deficiencies in the proposed approach to establishing class-wide loss. 

In contrast, the CAT held that the Off-Amazon Claim failed to meet the threshold for certification set out in Pro-Sys Consultants v Microsoft [2013] SCC 57 (Microsoft) on the basis that the PCR had not articulated a sufficiently coherent and factually grounded theory explaining how alleged price increases on Amazon's marketplace would have influenced pricing across wider retail channels. The expert methodology was therefore regarded as too speculative and insufficiently developed to justify certification.

In assessing the claims advanced by the class representative, the CAT also provided a clear synthesis of Microsoft, and its role in the certification test. The CAT reiterated that certification is not a merits assessment and should not become a mini-trial. Instead, the key question is whether the proposed methodology provides a plausible and workable (as opposed to “purely theoretical”) "blueprint" for resolving common issues and proving loss at trial ([50]-[51]). 

Key takeaways

Expert independence under the spotlight

The CAT gave detailed consideration to the role of Dr Chris Pike, who had been involved in developing the claim before being formally instructed as an expert witness.

The CAT accepted the Proposed Defendants’ suggestion that Dr Pike had a reputational interest in the success of the case which materially differed from that of an expert who is approached by a solicitor contemplating bringing a collective action. An expert approached by lawyers considering a claim can, if appropriate, conclude that a proposed theory of harm is unsustainable. By contrast, where the expert has helped develop the claim itself, reversing position may carry reputational consequences.

Those factors did not inevitably mean that an expert who had initiated proceedings cannot properly give evidence, but they underscore the need for that expert to be particularly mindful of impartiality and to be able to demonstrate it ([83]-[84).

Against that backdrop, the CAT concluded that parts of Dr Pike's third report were affected by his lack of objectivity, were defensive and adversarial, and crossed the line from expert analysis into advocacy on behalf of the class representative. While declining to exclude the evidence altogether, the CAT chose to treat Dr Pike’s evidence with “considerable caution” and attached limited weight to certain aspects of the evidence relating to the Off-Amazon claim ([93]-[94]).

The Proposed Defendants also argued that the entirety of Dr Pike’s evidence (including his evidence in relation to the On-Amazon Claim), should be given no weight, given the concerns regarding his independence. The CAT, however, disagreed, and found that Dr Pike’s evidence in relation to the On-Amazon Claim was supported by initial empirical analysis, which was not objected to by either Proposed Defendant.  

The CAT also made clear that any decision regarding whether Dr Pike should give further evidence, and the weight (if any) that ought to be attributed to any such evidence, is ultimately a case management decision for the Tribunal, following the approach taken in R (Factortame) v SS for Transport (No. 8) [2002] EWCA Civ 932. 

The decision therefore serves as a reminder that experts who have played a role in originating claims may face particular scrutiny regarding their independence and that experts must be vigilant in avoiding any appearance of advocacy in developing and presenting evidence. 

Distribution Matters

The judgment also reinforces the growing importance of distribution arrangements at the certification stage.

For the On-Amazon Claim, the proposed class representative relied on a purported direct-credit mechanism which was claimed to be able to use Apple and Amazon transaction data to identify class members and facilitate payments. The CAT considered that this provided a realistic prospect of achieving meaningful uptake, notwithstanding the modest value of individual claims. 

By contrast, distribution for the Off-Amazon Claim would have depended much more heavily on consumers actively submitting claims through an online process. The CAT was notably less confident that such a mechanism would generate significant participation ([161]-[167]).

While not determinative, the judgment continues the trend of the CAT placing increasing weight on whether damages can realistically reach class members in choosing whether to certify a claim. 

Funding arrangements and the legacy of Riefa

The CAT also examined whether costs incurred in the unsuccessful Riefa proceedings could be recovered from any damages obtained in the reconstituted action.

Under the original funding arrangements, the funder, insurers and advisers could potentially have recovered costs linked to the failed Riefa certification application from any future recovery achieved in the new proceedings.

In circumstances where the Riefa CPO application had failed in its entirety, resulting in final judgment in favour of the Proposed Defendants with consequential costs orders, the CAT concluded that it would be “manifestly unreasonable and inappropriate” for costs associated with an unsuccessful certification application to be transferred to a subsequent action and ultimately borne by class members. Certification was therefore made conditional on the removal of those provisions ([211]). 

The CAT made it clear that, notwithstanding its willingness to engage on this occasion, certification is not intended to be a negotiation with the Tribunal. Instead, parties are expected to carefully consider and put forward compliant funding arrangements from the outset rather than relying on the Tribunal to require amendments as a condition of approval. A failure to do so may result in a decision to refuse certification going forward ([234]). 

A developing benchmark for class representative remuneration 

The CAT also expanded on its recent comments in Waterside Class Limited v Mowi ASA & Ors [2026] CAT 32 regarding proposed class representative remuneration.

The Tribunal capped Mr Le Patourel's future remuneration at £170/hour and £750/day, subject to annual inflationary adjustments. In doing so, the CAT drew parallels between the role of a class representative with public sector fee-paid panel and non-executive roles (such as CMA panel members or Financial Ombudsman Service non-executive directors). 

Importantly, the CAT indicated that the £170 hourly rate reflected case-specific considerations, including work that Mr Le Patourel had already undertaken at a higher rate. Absent those factors, the Tribunal suggested that a rate of approximately £150/hour would likely have represented the upper end of what was reasonable ([272]-[275]).

The judgment makes it clear that class representative remuneration is now firmly subject to active judicial oversight, with the CAT setting a ceiling of what is considered reasonable and signalling an expectation that remuneration arrangements should generally be disclosed to class members through claim websites in the interests of transparency 

While the CAT emphasised that a future consultation may lead to further developments, the judgment provides a significant reference point, which may effectively limit class representative remuneration going forward.

Conclusion

The CAT's judgment marks another significant step in the development of the UK's collective proceedings regime.

Most immediately, it demonstrates that certification is not an all-or-nothing exercise. The CAT has certified one theory of harm while refusing another within the same application, allowing a narrower claim to proceed where it was supported by a sufficiently developed methodology and preliminary evidence.

More broadly, the judgment signals the CAT’s willingness to adopt a more demanding approach to certification. While the CAT reaffirmed that that proposed methodologies must be grounded in the facts of the case and supported by a credible route to establishing class-wide loss. The decision also highlights the increasing importance of practical distribution mechanisms, careful funding structures, demonstrable expert independence and proportionate class representative remuneration.