Dhan Kumar Limbu & Ors v Dyson Technology Limited & Ors

This action is brought by or on behalf of twenty-four Nepalese and Bangladeshi migrant workers in respect of working and living conditions in two Malaysian factories manufacturing components for the Dyson supply chain. The parties agree that it is appropriate to order that the claim should be tried in stages and that it will be necessary at some point to identify lead claimants. This judgment is principally concerned with whether the court should order split trials of liability and quantum or whether it is appropriate to order the trial of preliminary issues as to the defendants’ potential liability upon the pleaded facts, and the associated question of when lead claimants should be identified.

The two factories are in Johor Bahru, Malaysia and were operated by two associated Malaysian companies, ATA Industrial (M) Sdn Bhd and Jabco Filter System Sdn Bhd. The migrant workers allege that they were unlawfully trafficked to Malaysia and that they were subjected to conditions of forced labour, exploitative and abusive working and living conditions and, in some cases, detention, torture and beating in the course of their employment at the factories. These claims are brought against Dyson Technology Limited, an English company concerned with the invention, development, sale and service of domestic appliances and commercial hand dryers and batteries; Dyson Limited, an English company concerned with the sale and service of domestic appliances and commercial hand dryers; and Dyson Manufacturing Sdn Bhd, a Malaysian company concerned with the manufacture, sale and distribution of Dyson products.

The claimants argue that the Dyson companies controlled the working and living conditions at the factories through detailed contractual provisions with ATA and Jabco. Further, they argue that Dyson exerted a high degree of control over operations and conditions at the factories and that Dyson knew, or ought reasonably to have known, of the high risk that workers might be subjected to unlawful forced labour and denied their minimum legal employment rights. The claim is pleaded in negligence, false imprisonment, intimidation, assault, battery, and unjust enrichment:

The defendants deny any liability to the claimants. They specifically deny that they assumed any responsibility towards the migrant workers in the Malaysian factories and that they owed such workers any duty of care. Dyson put the claimants to proof of their alleged working and living conditions while denying actual knowledge of the same. They argue that the claimants wrongly seek to hold Dyson liable for the actions of independent third parties that the defendants neither owned nor controlled. Dyson assert that many of the matters alleged are outside their knowledge and therefore they put the claimants to proof. Dyson deny any joint liability for intentional torts committed by third parties. Further, Dyson deny any liability in unjust enrichment. Such claim is, Dyson argue, misconceived in Malaysian law. Finally, Dyson assert that a number of claims are time-barred.

Proceedings were issued on 27 May 2022. Dyson disputed jurisdiction arguing that Malaysia was the more appropriate forum. That issue was initially determined in Dyson’s favour but, by a judgment handed down on 13 December 2024, the Court of Appeal rejected the jurisdiction challenge and lifted the stay imposed by the deputy judge: see Limbu & Others v. Dyson Technology Ltd & Others [2023] EWHC 2592 (KB) and, on appeal, [2024] EWCA Civ 1564, [2025] I.C.R. 560. There was then some further delay while Dyson sought permission to appeal the issue of jurisdiction to the Supreme Court. Permission was refused on 6 May 2025 and, accordingly, this action now proceeds in this jurisdiction although it is common ground that the applicable law is Malaysian.

The claimants filed Amended Particulars of Claim on 22 July 2025, Dyson filed their defence on 12 September 2025 and the claimants filed a Reply on 24 October 2025.

I heard the first costs and case management conference in this case on 19 December 2025. While the parties agree a number of directions in principle, they take fundamentally different positions as to the appropriate issues to be tried at stage 1 and as to when the court should seek to identify lead claimants. There are also issues as to the appropriate expert evidence at this stage. Further, there are applications to strike out passages in the Amended Particulars of Claim and Reply, and a cross-application to re-amend the Particulars of Claim and amend the Reply. In addition, there is an issue as to whether the court should order early disclosure of certain key documents. Finally, there is the question of cost budgeting.

In managing this case, I have particular regard to the need to ensure that this litigation is dealt with justly and at proportionate cost. The following aspects of the overriding objective are particularly relevant in this case:

Before turning to the arguments, I should also record some uncertainty as to whether the court has all of the final parties before it:

Edward Craven KC, Piers Feltham, Thomas Fairclough and Joshua Jackson, who appear for the claimants, submit that the court should identify six lead claimants and try all issues of liability at stage 1. Mr Craven concedes that the court would have to determine whether the alleged breaches of duty caused loss while questions of the quantification of any such loss would be left over for a subsequent trial. Such trial would require expert evidence as to Malaysian law and, he asserts, forensic accountancy evidence in order to establish the claimants’ case as to unjust enrichment. The claimants’ proposal would, he argues, allow the court to make findings as to liability on the basis of actual findings of fact as to the working and living conditions in the two factories. Such a trial could, he submits, be accommodated within a 20-day listing from April 2027.

Against that, Mr Sloboda and Ms Srirangam submit that the court should try the following threshold issues of liability at stage 1:

Although it is not entirely clear from such formulation, Mr Sloboda asserts that these issues can be tried relatively conveniently without requiring the claimants to give evidence. Accordingly, he envisages that the negligence issues would also proceed on assumed facts. Thus, the defendants’ proposal is that these threshold issues of liability are determined on the assumption that the claimants will establish the fact of their employment, the exploitative and abusive working and living conditions, the commission of the intentional torts, and that the Dyson companies were enriched as alleged in each case in the Amended Particulars of Claim. Such trial could he submits, be accommodated within a 20-day listing from April 2027.

Mr Sloboda argues that the trial of threshold issues of liability at stage 1 without, at that stage, considering the merits of individual claims was the approach taken in two of the largest class actions currently before the High Court. He cites the approach advocated by the Court of Appeal, and ultimately taken by O’Farrell J, in Mariana, at [188], and the approach taken by the joint managing judges, Cockerill LJ and Constable J, in the Pan NOx Emissions Litigation as explained at [2024] EWHC 1728 (KB), at [10]-[11].

In arguing against the claimants’ proposal, Mr Sloboda cautions that the process of identifying lead claimants will be complex and time consuming. Further, he argues that lead claimants cannot be selected until the anticipated additional claimants have been joined and pleaded their cases. There would then be a need for individual Particulars of Claim, Defences and Replies. The proposal would also require the migrant workers to give evidence both at the liability trial and then again at the quantum trial.

Mr Sloboda argues that the claimants’ proposal to try liability issues in respect of unjust enrichment would require extensive review and disclosure of financial documents over a decade of trading. Further, the proposal would involve substantial lay evidence from the claimants, defence lay evidence, expert evidence as to Malaysian law, forensic accountancy evidence, and possibly medical evidence in respect of the issue of causation. The need to establish causation would, he submits, prevent there being a clean split between liability and quantum. Such a trial could not, he submits, be heard in 20 days and would require a 42-day listing.

In reply, Mr Craven cautions that Dyson’s approach would require the court to try the threshold liability issues on assumed facts. Citing Neuberger J’s well-known observations in Steele v. Steele [2001] C.P. Rep. 106, Mr Craven urges me to resist the siren song of determining preliminary issues on assumed facts that might later prove to be disputed by Dyson. He particularly identifies the problem of how the court can try Dyson’s alleged knowledge of facts that are, on their approach, to be assumed. Further, he submits that the defendants’ proposal seeks to deny the claimants a voice, delay by a matter of years the claimants’ opportunity to give their evidence, and set up a trial that can only finally dispose of proceedings in favour of Dyson.

Asked whether the claimants’ own proposals risk their electing to call a substantial number of claimants in support of the case advanced by the lead claimants, Mr Craven responded that their current intention was to call no more than ten witnesses as to conditions in the factories, namely six lead claimants and four additional witnesses.

I should also record that much of the parties’ pre-hearing correspondence and written submissions were devoted to seeking to establish that their opponents had changed their previous positions on these matters. Certainly, it is true that counsel representing Dyson before the Court of Appeal then anticipated that the court should first try the issue of liability on a lead claimant basis while the claimants’ position has also evolved from an earlier apparent acceptance that only certain “generic issues of liability” should be tried and the omission (which might have been no more than oversight) of any initial proposal to identify lead claimants. With limited success, I encouraged the parties not to focus on seeking to persuade me that their opponents had changed their position but upon the actual advantages and disadvantages of the final proposals now before the court. In fairness, Leigh Day was right in its letter of 18 November 2025 to suggest that it would be more productive to focus attention on the case management directions rather than rehash the parties’ different interpretations of past correspondence, and Mr Craven was right to anticipate in his written submissions that the court was unlikely to be interested with petty procedural sniping. For the avoidance of doubt, I am not assisted in determining the proper management of this case by tit-for-tat points as to whose position has changed. My focus is upon identifying the management of this case that will best allow me to deal with this case justly and at proportionate cost.

No group litigation order has been made or sought pursuant to r.19.22. Nevertheless, this action is already brought by twenty-four claimants and it is envisaged that that number may soon be well over one hundred. Even with twenty-four claimants, the position has already been reached where there are a significant number of claimants before the court with very similar claims in respect of working and living conditions at the two Malaysian factories. Each individual claim is likely to be relatively modest in value compared with the enormous cost of this litigation and it is likely that decisions on liability reached in respect of a number of lead claimants would be extremely valuable in allowing the parties to settle the claims brought by the remaining claimants. Accordingly, this case is in my judgment a paradigm example of a case where it would be convenient to try a representative sample of lead cases.

This analysis is not in dispute. The difference between the parties is not as to whether the court should try lead claims but when. The claimants argue that lead claimants should be identified immediately and that the court should order a split trial of the issue of liability in respect of the lead claims. They argue that lead claimants can readily be selected without delay and that any new claims should simply be stayed pending trial of the lead claims. As already explained, the defendants respond that it is not yet necessary or desirable to identify lead claimants and that the court should first try the threshold liability issues at stage 1 which can be tried without any evidence from the claimants.

No party applied to adjourn the case management conference pending greater clarity as to any possible application to join an additional 70-100 claimants or to consolidate proceedings should such new claimants issue a fresh claim. They were right not to do so; these proceedings have already been on foot for over 3½ years and there is no certainty as to if or when any such application might be made. The court can only manage the case that is currently before it and, accordingly, I must focus on the claim by these twenty-four claimants. Equally I do not seek to prejudge the approach that might be taken should a substantial raft of new claimants be joined in this action save to observe that one important consideration will be to ensure that the case is not blown off course.

In Lancaster v. Peacock [2020] EWHC 1231 (Ch), Fancourt J sitting with Master Kaye addressed the number of lead claimants that might be appropriate in a case. They explained, at [3]:

“It is not, of course, necessary to have very many sample claimants in order to decide common issues. The purpose of a broader selection of sample claimants, beyond what is needed to try the common issues, is to generate sufficiently broad guidance for the likely disposal of all the other claims, whose particular facts will vary, while at the same time not overcomplicating or encumbering or significantly adding to the cost of the trial.”

For reasons that I explain more fully below, I consider that lead claimants should be identified at this stage. Accordingly, I shall invite submissions at the adjourned case management conference as to the proper selection of lead claimants but venture the following initial views lest they assist the parties in understanding my current thinking:

It is common ground that the court should not order a full trial of all issues even on a lead claimant basis. As discussed above, the battleground is as to whether the court should order split trials of the issues of liability and quantum, or whether the court should first try discrete issues as to the defendants’ potential liability as preliminary issues.

In the course of argument there was some discussion as to whether there was any real difference between the trial of a preliminary issue and a split trial. In complex personal injury and clinical negligence litigation, issues of liability are routinely tried separately from issues of quantum. While often referred to as a split trial, the King’s Bench Guide makes the point that in such cases the court is in reality ordering the trial of liability as a preliminary issue. The precise terminology may not matter but there is a stark difference between a conventional split trial of the issue of liability and a trial of preliminary issues in the sense explained by David Steel J in McLoughlin v. Jones [2001] EWCA Civ 1743, [2002] Q.B. 1312, at [66], as the trial of a potentially decisive question, usually of law, that can be decided on the basis of a schedule of agreed or assumed facts. That distinction was rightly recognised by Peter Macdonald Eggers KC sitting as a deputy judge in Jinxin Inc. v. Aser Media Pte Ltd [2022] EWHC 2431 (Comm), at [20]:

“On some occasions, split trials and trials of preliminary issues are terms which are used interchangeably. This is for good reason: because they often amount to the same thing. However, the archetypical preliminary issues trial is one which requires the determination of one or a few issues (often, but not always, an issue of law) whose resolution depends on the court drawing certain assumptions, which generally occupies the court for a short period of time, and whose outcome has the real potential to save the parties and the court of the need for a substantially longer trial. By comparison, a split trial may involve all of the issues being determined at two or more separate hearings, each concerned with a substantial number of issues both of law and fact, and yet the first hearing has again the real potential to avoid the need for a subsequent hearing or subsequent hearings. Very often, where trials are split, the first trial need not make any assumptions of fact. This may be where the distinction between split trials and preliminary issues trials resides. In any case, it is the potential to save the court and the parties from having to deal, at a subsequent hearing, with issues which become irrelevant depending on the decision reached by the court at the first trial which is common to both preliminary issues trials and split trials.”

Trials of preliminary issues in the sense identified in McLoughlin have their value but can, as Lord Scarman observed in Tilling v. Whiteman [1980] A.C. 1, at p.25, all too often prove to be treacherous short cuts whose price can be delay, anxiety and expense. Tilling was a simple county court possession claim in which, rather than embark on a straightforward exercise of hearing the evidence, the judge decided the preliminary issue of whether on assumed facts the landlords would be entitled to possession. The judge decided that issue against the landlords and dismissed the possession claim. Two appeals later, the House of Lords found that the judge had wrongly decided the preliminary issue and, since there were no findings of fact, there was no option other than to remit the case for trial to the county court. Deprecating the course that proceedings had taken, Lord Wilberforce observed, at pp.17-18:

“So the case has reached this House on hypothetical facts, the correctness of which remain to be tried. I, with others of your Lordships, have often protested against the practice of allowing preliminary points to be taken, since this course frequently adds to the difficulties of courts of appeal and tends to increase the cost and time of legal proceedings. If this practice cannot be confined to cases where the facts are complicated and the legal issue short and easily decided, cases outside this guiding principle should at least be exceptional.”

Lord Neuberger often warned of the dangers inherent in ordering the trial of preliminary issues:

In Woodland v. Swimming Teachers Association [2013] UKSC 66, [2014] A.C. 537, a school pupil suffered severe brain damage during a swimming lesson that had been arranged by her school but took place at a pool operated by a third party. She sued the Swimming Teachers Association, the contractor that employed the swimming teacher and the lifeguard, the lifeguard herself, the education authority that ran her school, and a further local authority that operated the pool. At first instance, the judge tried as a preliminary issue the question of whether the education authority owed a school pupil a non-delegable duty to secure that reasonable care was taken of her during the school day. Finding that no such duty was owed, the judge struck out that aspect of the claim against the education authority. On a second appeal, the Supreme Court held that the education authority could in law be liable for injury caused by the negligence of an independent contractor to whom it had delegated its education function and control over the pupil during the school day. Accordingly, the case had to be remitted. Lord Sumption commented, at [2]:

“The appeal provides a useful occasion for reviewing the law on what have been called ‘non-delegable duties of care’. But it must be very doubtful whether deciding such a point on the pleadings was really in the interests of these parties or of the efficient conduct of their litigation. The pleadings are unsatisfactory. There are no findings of fact and almost everything is disputed. A decision of the point presently before us will not be decisive of the litigation either way because there are other bases of claim independent of it. The point has taken more than two years to reach this stage, during which, if the allegation had been allowed to go to trial, it would almost certainly have been decided by now. As it is, regardless of the outcome of this appeal it will now have to go back to the High Court to find the relevant facts.”

To similar effect, Lord Hope observed in SCA Packaging Ltd v. Boyle [2009] UKHL 37 that the power of a tribunal to order the hearing of a preliminary issue should be “exercised with caution and resorted to only sparingly”. He endorsed Lindsay J’s observation in an earlier case that the essential criterion was whether there was a succinct knockout point that was capable of being decided after only a relatively short hearing. That, Lord Hope added, was unlikely to be the case where the preliminary issue could not be entirely divorced from the merits of the case or where it would require consideration of a substantial body of evidence.

In my judgment, the defendants’ proposal has the superficial attraction of cutting to the chase of determining what might well prove to be the central issues in this case as to the defendants’ control and auditing of the factories and the extent of their knowledge of the treatment of the migrant workers. Such issues will be central to the question of whether the Dyson companies owed a duty of care to the claimants and, if they did, the question of liability in negligence. Further, such issues will be important in determining whether the Dyson companies are jointly liable for the alleged intentional torts and the success of the claim in unjust enrichment.

There are, however, a number of sound reasons why, in my judgment, the court should not be distracted by the siren song extolling the potential virtues of the defendants’ proposed threshold issues trial:

I therefore consider that the defendants’ proposal would not allow the court to deal with this case justly and at proportionate cost. In particular, acceding to Dyson’s proposals would impede the fair resolution of issues on assumed facts both at first instance and on appeal, and would breach my duty to ensure that this litigation, which has already been delayed by the jurisdiction challenge, is now progressed promptly.

I acknowledge that the stage 1 trial in Mariana did not focus on individual claims but rather upon questions as to the defendants’ liability for the collapse of the Fundao dam in Brazil and the widespread environmental damage that was caused. The issues tried at stage 1 were broadly whether the defendants were strictly liable under Brazilian law for environmental damage and whether they were liable for fault-based liability under the Brazilian Civil Code or the Brazilian Corporate Law. Anticipating such an approach, the Court of Appeal had observed in dismissing the jurisdiction challenge, at [188]:

“Significantly, by [the stage of the case management conference] the parties, represented by sophisticated litigators experienced in the field, would have been obliged fully to co-operate in putting forward case management proposals. Those could have included, for example, a proposal that the issue of indirect polluter strict liability be tried preliminarily, or that the control basis of such liability be determined first. A trial of such a preliminary issue would not appear to require the involvement of any evidence from any individual claimants; rather it would turn on the defendants’ knowledge, control and conduct, and the scope of the dispute on these aspects of the claim would be apparent from the defendants’ responsive pleading to the [Master Particulars of Claim]. Were liability to be established, generic causation and quantum issues could be identified and tried. The outcomes of quantum trials on individual selected test cases could be used to inform settlement of individual cases across the board.”

Thus, the stage 1 liability issues in Mariana did not turn on anything that the claimants could say in evidence but upon generic questions as to the proper application of the strict liability rules and the defendants’ liability for the fault-based liability of the impecunious owner and operator. It was only if liability was established on one or more of those bases that it would be necessary to consider whether individual claimants had suffered loss and damage by reason of the collapse of the dam.

Likewise, I acknowledge that threshold liability issues were ordered to be tried in the Pan-NOx litigation without identifying lead claimants or trying any individual claims. In that case, the claimants alleged that car manufacturers had sold diesel cars that did not comply with applicable regulatory regimes in that they contained software that cheated or subverted their true level of emissions. The court ordered that the first trial should determine whether it was bound by findings of the German Federal Motor Transport Authority. A second trial was described as the main evidential phase and would determine whether the cars contained defeat devices, whether that was justified, legal issues as to the construction of the Emissions Regulation, and whether the supply of cars with defeat devices breached EU law and/or contracts of sale. These were generic issues that did not require the identification of or involvement of lead claimants.

I do not find either Mariana or Pan NOx to be helpful in formulating the proper approach in this case. Mariana concerns a single catastrophic event that had profound consequences for hundreds of thousands of people and businesses. There is no doubt that the dam collapsed causing loss of life and widespread environmental and physical damage. There may subsequently be an issue as to whether individual claimants suffered loss and damage by reason of the disaster, but the claimants have little if anything to say about liability. There are potentially millions of claimants in Pan NOx, being those who bought diesel cars fitted with defeat devices. Technical witnesses can assist with how the cars were mass produced and there was unlikely to be anything useful to the issue of liability that any individual claimant would be able to say.

This is, however, a very different case. Here, the narrative starts with the individual claimants. The core factual assertions are that these migrant workers were subjected to exploitative and abusive working and living conditions while employed in the Dyson supply chain. It is that alleged ill-treatment that lies at the heart of the case and in respect of which it is alleged that the defendants owed and breached a duty of care, and that they were jointly liable for the intentional torts committed by those who managed the factories. Importantly, the alleged ill-treatment gives content to the allegation that the defendants knew and controlled the true state of affairs at the factories. In my judgment, resolution of the issue of liability in respect of a representative number of lead claimants will allow the court to consider the key issues as to Dyson’s liability, if any, for such conditions on the basis of actual findings of fact rather than on the uncertain basis of assumed facts.

With the exception of the claim in unjust enrichment, it is not, in my judgment, particularly complicated to try liability issues in respect of lead claimants. Issues as to the alleged mistreatment of migrant workers will largely turn on witness evidence from those who worked in, managed, visited, or audited the factories as to working and living conditions; any photographic or documentary evidence as to these matters and in particular any inspection or audit reports; and any records of hours worked by individual workers, the wages paid, the deductions made from their wages, and the holidays taken. Issues as to whether such working and living conditions were unlawful will turn upon those findings of fact and the expert evidence of Malaysian law. Further, issues as to Dyson’s liability, if any, for such working and living conditions will turn on those findings of fact; evidence as to the companies’ control of the operations in the factories; evidence as to their knowledge of such conditions; and further expert evidence upon interesting questions of Malaysian law as to Dyson’s alleged liability for its failure properly to control the management of its supply chain.

No one suggests that quantum issues as to the extent of any injury, loss and damage should be tried at stage 1. I agree and there is of course nothing exceptional about split trials of liability and quantum.

There was some debate in the course of argument as to the issue of causation. While I was not addressed on any differences between English and Malaysian law on this issue, it may be that it is necessary to prove some loss to establish liability for some causes of action while other torts are actionable without proof of loss. Certainly that is the position in England & Wales where, for example, a negligence claim requires proof of loss whereas the torts of false imprisonment, assault and battery are actionable without proof of loss. That was therefore also the law of Peninsular Malaysia on 7 April 1956 since the Civil Law Act 1956 adopted the English common law in preparation for independence the following year. Whether it remains the law of Peninsular Malaysia will be an issue – should it be in dispute – for the experts in Malaysian law.

The fact that – depending on Malaysian law on this point - some evidence may be required to establish that at least some injury, loss or damage was caused but not to quantify the same is not a novel problem. For example, the fact that a personal injury claimant in this jurisdiction cannot establish liability without establishing that the defendant’s negligence or breach of statutory duty caused at least some injury, loss or damage does not prevent the courts from hearing split trials of liability and quantum. Nor does it usually require any medical evidence at the liability trial, it being common ground – or at least easily established without medical evidence - that the claimant suffered at least some injury, loss or damage by reason of the alleged negligence or breach of statutory duty. Likewise here, if a claimant establishes that the defendants breached any duty of care owed to him or her and that but for such breach of duty the claimant would not have been:

That said, the parties should co-operate in keeping the issue of causation under active review. If a real causation issue is identified in relation to any particular cause of action pursued by a claimant which under Malaysian law is arguably only actionable on proof of loss and upon which that claimant might require medical or other expert evidence at the liability trial then the parties should consider whether that is a reason to remove that person from the list of lead claimants. Alternatively, they should consider whether they might require limited expert evidence on that issue. The court will not be impressed by gamesmanship on this issue but seeks the parties’ co-operation and assistance in ensuring that the necessary liability issues are fairly tried without simply defaulting to the safety-first and expensive approach of needlessly relying on medical or other expert evidence in each case at the liability stage. That calls for co-operation and the early identification of a problem, and neither an opportunistic defence ambush at trial on the basis that a particular claim fails for want of expert evidence on causation nor a failure by the claimants’ legal team to engage with the issue.

I have given anxious consideration to whether the claim in unjust enrichment should be determined at the liability trial. I acknowledge that, at the time of the jurisdiction appeal, it was assumed that all liability issues would be heard together. Further, I acknowledge that its exclusion would be untidy in that issues of liability in unjust enrichment would then fall to be determined at the quantum hearing. Against that, making a finding as to the unjust enrichment claim is not essential to determining what happened in the factories or the extent to which the defendants are liable either in negligence or jointly in respect of the intentional torts. In fact, the unjust enrichment claim is parasitic upon establishing the alleged exploitative and abusive working and living conditions. Indeed, the core premise (pleaded at paragraph 94.1 of the Amended Particulars of Claim) is that Dyson obtained the manufacture of parts and products at a cost significantly below the true and fair market price that they would have had to pay had Dyson contracted with factories that respected the human and employment rights of its workers. Such enrichment is alleged to have been at the expense of the claimants on the basis of the alleged exploitative and abusive working conditions and it is asserted that the claimants’ high degree of vulnerability and the extent and egregiousness of their exploitation set against the size of the profits made by the defendants require the court to grant the restitutionary remedy of unjust enrichment (paragraphs 95 and 97).

In my judgment, the issues of liability for and the assessment of the extent of any unjust enrichment cannot be safely or conveniently split from each other. Further, it would be disproportionate and inconvenient to give permission to the parties to rely on forensic accountancy evidence in respect of the liability aspect of unjust enrichment before trying issues as to the working and living conditions in the factories and of liability in respect of the negligence and intentional tort claims. Disproportionate in that it would require expensive forensic accountancy evidence to be prepared and given twice: first at the liability trial and then – should liability be established - the experts would have to take into account the court’s findings in the first trial, prepare further reports and ultimately give expert evidence as to the quantum of the unjust enrichment for the purposes of the quantum trial. And inconvenient since it is difficult to see how the forensic accountants could easily identify and assess the alleged enrichment without first being guided by the court’s actual findings of fact as to the working and living conditions in the factories and Dyson’s knowledge or control of such conditions.

For these reasons:

At the hearing on 19 December, I granted the parties permission to rely on expert evidence as to Malaysian law. While they have already agreed draft directions for the orderly exchange of reports, the administration of written questions, and the preparation of joint reports, I do not grant permission at this stage to call oral expert evidence at the liability trial. That issue must be considered at the pre-trial review once it is clear what the disputes of Malaysian law are.

At this stage, no other expert evidence is necessary for the liability trial.

By an application made on 21 November 2025, the defendants applied for an order striking out paragraphs 98-99 of the Amended Particulars of Claim and over 200 paragraphs and sub-paragraphs of the Reply. By a cross-application dated 11 December 2025, the claimants sought permission:

There is no dispute that large swathes of the Reply should come out:

Such dispute is indeed arid, but the starting point is that the Reply was defective and there was clearly no justification for a 66-page Reply in this case:

In my judgment, the claimants’ Reply breached each of these cardinal principles. As a direct result, it was substantially longer than it should have been. While often the court will not strike out a statement of case simply because it pleads some irrelevant issues or generates some untidiness in the pleadings, there is a distinction between a case in which an occasional lapse in pleading practice causes the type of untidiness tolerated by Jackson J in Atos Consulting Ltd v. Avis Europe plc [2005] EWHC 982 (TCC), [2005] CP Rep 43, at [18], and cases such as the present where the wholesale breach of the rules would otherwise leave in place a 66-page Reply that would serve only to obfuscate whether there was anything of value in the document. In my judgment, this Reply is likely to obstruct the just disposal of the proceedings and the offending passages were therefore liable to be struck out pursuant to r.3.4(2)(b).

When faced with a meritorious strike-out application, it was somewhat tactical to seek to withdraw the impugned parts of the Reply by amendment rather than taking the more straightforward path of consenting to the strike-out application. In any event, the strike-out application was first in time and, insofar as it matters, I strike out the passages that the parties agree should be removed from the Reply.

The nine paragraphs still in dispute consist variously of commentary, repetition and statements of law. They serve no purpose and should not have been pleaded in the Reply. Had these nine paragraphs been the only objectionable passages in the Reply, I might well have been persuaded that they introduced some untidiness but did not warrant an application to strike out. That was not, however, the position and I also strike out these nine paragraphs.

The position is different in respect of the Amended Particulars of Claim. Paragraphs 98-99 should not have been pleaded but would not of themselves have justified a strike-out application. Accordingly, I would allow the claimants to delete those two paragraphs by way of re-amendment.

It is, however, perfectly obvious that this issue has become all about costs and that all parties are jockeying for forensic advantage:

Having substantially succeeded in the strike-out application, the general rule is of course that the defendants should be awarded their costs. The court is, however, required to have regard to all the circumstances of the case including conduct and specifically whether it was reasonable for a party to raise, pursue or contest a particular issue, and the manner in which the parties pursued or defended the issue: r.44.2(4)-(5). Applying these principles, I conclude:

This was therefore a meritorious and reasonable application that was pursued somewhat aggressively and without proper regard to the parties’ duty under r.1.3. In my judgment, the appropriate costs order is that the claimants should pay 50% of the defendants’ costs of the strike-out application. There are no proper grounds for awarding costs on the indemnity basis and accordingly I will assess costs on the standard basis.

The defendants have provided a schedule of costs in respect of the application totalling £61,366.03. I accept Mr Craven’s argument that to spend over £60,000 on this issue was both unreasonable and disproportionate. The application has been fought without any proper regard to the actual importance of the issue or the value of these claims. While Mr Sloboda’s advice was properly sought, the issue should then have been delegated to junior counsel. Equally, the defendants were entitled to the advice of the partner responsible for this litigation but not to their claim for over 18 hours of her time working on the documents. I accept that it is somewhat involved to identify the precise targets for strike out amidst a 66-page pleading, but this application did not remotely justify over 100 hours’ work on the documents (including over 12 hours’ work on costs) and over 8 hours’ attendance on counsel. Indeed, the work required to argue this application substantially reduced as soon as it became clear that the claimants would not seek to defend the vast majority of the impugned paragraphs.

Furthermore, counsel’s fees are claimed as a proportion of the brief fees for the entire costs and case management conference. I do not accept either the premise that the reasonable and proportionate total spend on counsel’s brief fees for the case management conference was anything like £105,000 nor the suggestion that the remaining issues of whether (1) the offending paragraphs should be struck out or withdrawn by amendment, and (2) the disputed nine paragraphs should be struck out, justified anything approaching 20% of such brief fees.

When assessing costs on the standard basis, r.44.3 requires the court to resolve any dispute as to whether costs were reasonably and proportionately incurred or whether they were reasonable and proportionate in amount in favour of the paying party. In my judgment, the reasonable and proportionate costs of this application were no more than £12,000. Accordingly, I award the defendants the sum of £6,000, being 50% of such costs.

Let the parties take note. As I have already observed, it is particularly important in group litigation of this sort that the parties should comply with their duty under r.1.3 to help the court in furthering the overriding objective of dealing with the case justly and at proportionate cost. That calls for co-operation rather than opportunism, aggression or – on the other side – tactical manoeuvring and burying one’s head in the sand. In this instance, the defendants were right to criticise the drafting of the Reply and the claimants would have been better advised to have consented to the order sought and move on rather than make their cross-application in the hope of deflecting the court’s attention to the reality of the position or insist that certain paragraphs should not be struck out at all when in fact counsel was not prepared to advance any positive argument for their retention. Although largely successful in their application, the defendants have, however, pursued this matter aggressively and without proper regard to their duty to help the court in ensuring the just and proportionate resolution of this claim. They have in consequence recovered only a fraction of the unreasonable and disproportionate costs that they spent on this issue.

A separate issue arises as to the costs of the Reply. The defendants invite me to direct that in any event 50% of the costs of the Reply should be disallowed. It is undoubtedly right that some part of the costs of drafting the Reply will have been unreasonably incurred on pleading the many paragraphs and sub-paragraphs that should never have been in the Reply. That is a perfectly proper point for the costs judge to take into account on detailed assessment, but I agree with Mr Craven that I am not in a position to calculate the extent of the wasted costs incurred. Accordingly, I leave this issue for detailed assessment in due course should that be required and I do not tie the hands of the costs judge by making the direction sought by the defendants.

The claimants plead a claim for aggravated damages. Currently, they rely on three matters: Dyson’s alleged persistent and unjustified failure to prevent the claimants from being subjected to unlawful forced labour; Dyson’s alleged persistent failure to prevent the claimants from being exposed to an unacceptably high risk of serious personal injury; and Dyson’s alleged continued failure to apologise. By their application, the claimants seek to re-amend paragraph 124 of their Particulars of Claim to plead two additional particulars:

“124.4

The fact that in 2022 the Defendants issued a challenge to the jurisdiction of the English court which:

(a)

was calculated to impede the Claimants’ ability to pursue and obtain redress from the Defendants in respect of the severe exploitation and mistreatment which the Claimants had been subjected to; and

(b)

had the consequence (which was entirely foreseeable to the Defendants) of delaying the Claimants’ access to a remedy in this jurisdiction by a period of more than two years, in circumstances where the Defendants knew that the Claimants were in a position of exceptional vulnerability and hardship and that such delay would inevitably perpetuate and exacerbate that hardship.

124.5

The fact that the First and Second Defendants brought and pursued for more than two years libel proceedings against Channel Four Television Corporation and Independent Television News Limited in 2022, which were calculated to prevent legitimate journalistic reporting about the exploitation and mistreatment which the Claimants had been subjected to at the Factory Facilities and Accommodation.”

Resisting these amendments, Mr Sloboda argues that the new case is hopeless. He observes that the jurisdiction challenge was plainly arguable and indeed that Dyson succeeded at first instance. Equally, he argues that the libel claim was arguable even if it was subsequently discontinued. He argues that the claimants have failed to establish a real prospect of success upon these amendments and complains that it is entirely unclear how Dyson can respond to the amendments without waiving privilege in its legal advice.

Citing Commissioner for Police for the Metropolis v. Shaw [2012] I.C.R. 464, Mr Craven responds that it is settled law that the way in which litigation is pursued can give rise to a claim in aggravated damages. He submits that it is properly arguable that the jurisdiction challenge was calculated to impede these claims. Why else, he asks rhetorically, was jurisdiction fought so doggedly?

Given that the Court of Appeal found that these claims are primarily brought against English companies; that the Malaysian Dyson company was a group company for whom the litigation would be co-ordinated and conducted from England by English officers and employees; that Dyson were tough and difficult opponents who would take every step legitimately available to them to defeat the claims; and that the effects of declining jurisdiction included that the claimants:

The story of the libel proceedings is curious. The claim was issued against two national broadcasters and it is plainly arguable that such proceedings were at least in part designed to prevent journalists from publishing further reports about the alleged exploitation and mistreatment of the claimants. As Popplewell LJ remarked in the jurisdiction appeal, the libel proceedings were discontinued after the broadcasters pleaded a truth defence and shortly after Warby LJ gave permission for the libel defence to be relied on in the appeal. Popplewell LJ added, at [43]:

“No reason has been given by Dyson UK for such discontinuance. The timing and history is such that a cynical observer might infer that this was in whole or in part to obtain tactical advantage in this appeal. Mr Gibson suggests, however, that there are a number of legitimate reasons why [Dyson] might now have decided to discontinue which are unrelated to the effect it might have on this appeal, and I will proceed on that basis.”

As to privilege, the issue raised by these re-amendments is whether the defendants’ actions in issuing the jurisdiction challenge were calculated to impede the claims and whether the libel proceedings were calculated to prevent legitimate journalistic reporting. These are not questions as to the merits of the jurisdiction challenge or the libel proceedings but as to the defendants’ intentions in deciding to challenge jurisdiction and in deciding to issue libel proceedings. It is not beyond the defendants’ very experienced legal team to chart a careful course in responding to such allegations without waiving privilege.

In my judgment, the claimants succeed in overcoming the low hurdle of establishing that these re-amendments have a real, as opposed to fanciful, prospect of success. I therefore allow the re-amendments to paragraph 124 of the Particulars of Claim.

There was insufficient time to hear the argument on the capacity re-amendment on 19 December 2025 and, in any event, the defendants sought time to respond to that part of the application. Accordingly, I gave directions and that aspect of the claimants’ application will be heard at the adjourned case management conference if it remains in dispute.

The draft directions envisage disclosure being given by 11 June 2026. The claimants seek, however, early disclosure of five categories of documents that were referred to in the defamation proceedings:

By Slaughter and May’s letter dated 23 October 2025, the defendants confirmed that these documents would fall within their standard disclosure although they reserved the right to withhold inspection on the basis of privilege and to redact confidential information that was not relevant to this case.

Mr Craven submits that these five classes of documents are clearly highly relevant and very important. Acknowledging that the only question is the timing of disclosure, he argues that early inspection of key documents will help in facilitating an effective disclosure process and may lead to further investigation of factual issues or amendment of the pleaded case. It is, he contends, a targeted and proportionate request for disclosure of documents that Dyson will already have identified. It would, he submits, be consistent with Dyson’s duty of co-operation for it now to give this early disclosure.

In response, Mr Sloboda asks the rhetorical question as to why it matters whether the claimants get these documents in March or June. There is, he argues, no justification for early disclosure and this is not a case where the claimants have been unable to plead their case.

Generally, the court does not order specific disclosure until the parties have first given disclosure. Orders for early specific disclosure are not routine but may be made where, taking into account the overriding objective and the respective consequences of making or not making the order, the applicant demonstrates a proper basis for early disclosure: Bullring Ltd v. Laing O’Rourke Midlands [2016] EWHC 3092 (TCC), per Coulson J, at [20]. Coulson J added that for a proper basis to be identified, there had to be something important or significant which could be achieved by ordering early disclosure. There therefore had to be a good reason for ordering early disclosure although the judge rejected the suggestion that something out of the ordinary was required as “possibly overstating the position”.

In Cavallari v. Mercedes-Benz Group AG [2023] EWHC 1888 (KB), [2024] R.T.R. 1, Fraser J made clear, at [31], that an applicant’s inability to plead its case without early disclosure is a powerful factor in favour of such an order but that it is not an essential prerequisite. Cavallari is group litigation. Fraser J observed, at [33]-[36]:

“33.

In my judgment, although group litigation is of course governed by the CPR generally and also must take account of the overriding objective in the CPR, there are differences in terms of scale that mean when the general principles to disclosure – and in particular early disclosure – are applied, a different outcome might occur in respect of this subject. This is because in group litigation it is more likely that if a particular discrete document is known to exist, and to be directly relevant to the issues (regardless of whether that document is itself admissible as proof of its findings and conclusions, which the [Korean Fair Trade Commission (‘KFTC’) Decision] is not, as has been explained above) it would be more usual to order early disclosure of it, than if the litigation were more conventional involving very few parties. In group litigation such as this, I struggle to see that disclosure of some of these documents ought to be delayed merely because pleadings have not closed. I am not for a moment suggesting that early disclosure will more readily be ordered in group litigation; such orders will be relatively rare. But the ‘something important or significant’ in group litigation may more readily be satisfied in group litigation than otherwise.

34.

This is for two main reasons. Firstly, early disclosure of a document such as the unredacted KFTC Decision itself may, and in this case probably will, assist the parties in refining the issues between them in the group litigation generally. This assistance to the parties will also help to inform the court as to the direction of the group litigation in terms of what issues can usefully be resolved, when, and in which order. Case management is important in most, if not all, complex cases, but in group litigation it is even more important, given the nature, scale and duration of such cases …

35.

The second reason is that co-operation by the parties, required in any event of all parties under CPR Part 1, is even more important in group litigation. It is hardly co-operative for a party with a directly relevant document, which I find the KFTC to be, to take the position that in principle it is disclosable, but to resist that on the basis of ‘not yet’.

36.

Early disclosure in this case also assists in correcting what Mr de la Mare correctly describes as the information asymmetry between the parties. There is sufficient information in the public domain concerning the diesel emissions landscape for these many hundreds of thousands of claimants to consider, or suspect, they have a claim, and for the Generic Particulars of Claim to be drafted. However, the sooner their legal advisers are aware of the full content of the KFTC Decision the better, as this will help those advisers realise either their case is weaker than they thought, stronger, or perhaps about the same. Such detail can only helpfully advance the group litigation at an early stage.”

In Alame, Bean LJ observed, at [102], that in a case where there is both a substantial inequality of arms and asymmetry of information, all case management decisions should be informed by the overriding objective and, in particular, the court’s obligation to ensure so far as reasonably practicable that the parties are on an equal footing and can participate fully in proceedings.

Here, the claimants have been able to plead their case. They know better than anyone the actual working and living conditions at the factories, but there is an asymmetry of information between the parties as to what was reported to and known by Dyson. Quoting Lord Briggs’ own earlier observation in another case, Popplewell LJ commented in the jurisdiction appeal, at [57], that it is highly likely that there will be amendments given that, in a case of this kind, the importance of disclosure of documents internal to Dyson is “blindingly obvious”.

Dyson has already incurred costs of over £540,000 on disclosure in this litigation. Within that level of expenditure, it is inconceivable that the defendants have not already identified and considered the key documents now sought, especially since such documents were pleaded in the libel case. I do not therefore accept that it would be onerous now to require Dyson to disclose a small number of key documents. Dyson’s position is, however, protected since I do not order a specific search for documents. Rather, Dyson should co-operate by disclosing such of these apparently key documents as have already been found.

Balanced against that minor inconvenience, the problem with amendments made after disclosure or, even later after the exchange of witness evidence, is that they require the parties to reconsider their disclosure and witness evidence in light of the amended allegations. That can cause additional cost and prejudice the ability of the parties to comply with the procedural timetable. While not routine, I accept that there is therefore something important that can be achieved by a strictly limited order for early specific disclosure of a small number of known key documents that are likely to reveal the defendants’ knowledge or control of working and living conditions at the factories and whether Dyson was requiring overtime to be worked on rest days despite any such knowledge. There is therefore a proper basis for the order sought which is, in my judgment, in accordance with the overriding objective of dealing with this case justly and at proportionate cost in that it:

The order obtained does, however, come with responsibilities. Having asked for and obtained an order for limited early disclosure, the claimants would breach their own obligations under r.1.3 if they fail to consider such documents and formulate any necessary consequential re-amendments with expedition.

I therefore order the defendants to give early specific disclosure of known documents in each of the categories identified at paragraph 71 above.

Having considered the parties’ draft directions and their rival submissions, I accept the claimants’ argument that the liability trial that I have ordered will be ready for hearing by 15 April 2027. Now that the scope and approximate date of the trial have been resolved, it is necessary to finalise the directions in order to achieve the just and proportionate trial of the issue of liability on that timetable. In this area, there has been considerable co-operation and substantial progress has already been made. I now invite the parties to seek to agree the remaining directions or at least to narrow any remaining areas of dispute in advance of the adjourned case management hearing.

I particularly invite the parties to consider whether disclosure relevant to the unjust enrichment claim should be given together with disclosure on other liability issues (thereby avoiding the undesirability of argument as to what, if any, financial disclosure is in any event relevant to the liability trial and of substantial further time being sought for financial disclosure after the liability trial) or delayed until judgment has been given on the liability trial (thereby saving costs of disclosure that might ultimately not be needed).

The argument on other issues on 19 December 2025 meant that there was insufficient time to embark on costs management. It was not an exercise to be rushed given the total budgets for incurred and anticipated future expenditure of £5.2 million for the claimants and £7.5 million for the defendants. Furthermore, it would have been neither just nor sensible to have embarked on budgeting without first settling the issue of exactly what was to be tried. As I observed at the hearing in December, it will now be necessary for the parties to review their budgets in light of the case management decisions that I have taken and the issue of costs management will have to be determined at the adjourned case management conference.

At the December hearing, I gave notice that I was very concerned by the enormous incurred and budgeted expenditure in this case and that the parties should not expect their projected costs – where they are not agreed – to be approved in anything like the sums estimated in their budgets. I accept that those observations were very much a matter of first impression and that my comments were made before hearing argument on the budgets. The parties may seek to persuade me otherwise but, nevertheless, I invite the parties to co-operate in seeking to identify tangible cost savings in their final budgets.