Trappit SA & Ors v GBT Travel Services UK Limited

The claimants have discontinued their claim against the defendant. They have applied by application notice dated 24 July 2023 to vary the default position under CPR 38.6(1) so that the usual costs rule does not apply, arguing that the defendant’s conduct has been such that the proceedings have been unnecessarily protracted. The defendant resists this and asks that the court make an order that the claimants pay the defendant’s costs on an indemnity basis.

This claim has been subject to a two-tier confidentiality order. There are 3 lever arch bundles of documents before me. The key witness statements are:

Both solicitors have conduct of the claim for their respective clients and have an in-depth knowledge of the claim having made previous witness statements in respect of the proceedings.

A claimant may discontinue all or part of a claim against one or more defendants at any time: CPR 38.2(1). Notice of discontinuance takes effect on the date when the notice of discontinuance is served on the defendant and that is when the claim is brought to an end against the defendant: CPR 38.5(1)(2).

Pursuant to CPR 38.6,

“(1)

Unless the court orders otherwise, a claimant who discontinues is liable for the costs which a defendant against whom the claimant discontinues incurred on or before the date on which notice of discontinuance was served on the defendant.”

This is also consistent with the general rule set out in CPR 44.2(2)(a) that the unsuccessful party will be ordered to pay the costs of the successful party.

As the author of Civil Costs 6^th Ed comments at paragraph 2-030,

“ Under CPR r.38.6 the court has to be persuaded that it is just to depart from the normal rule that the discontinuing party should bear the other party’s costs. The burden of satisfying the court that some other consequence should follow, rests on the party discontinuing. Justice would normally lead to the conclusion that a defendant who defended himself at substantial expense against a claimant who changed his mind in the middle of the action for no good reason, other than he had re-evaluated the factors that had remained unchanged, should be compensated for his costs.”

So, the presumption is that the defendant will be awarded their costs up to the date of service of the discontinuance notice and the burden is therefore on the claimant to show that there is a good reason for that presumption to be disapplied. The notes to the White Book at 44.9.3 put the above quote in a similar way,

“It is clear from the terms of r. 38.6(1) and from the authorities that a claimant who seeks to persuade the court to depart from the normal position has to provide cogent reasons for doing so, and is unlikely to satisfy that requirement, save in unusual circumstances. A claimant who commenced proceedings takes upon himself the risk of the litigation. If he succeeds he can expect to recover his costs, but if he fails, or abandons the claim, at whatever stage in the process, it is normally unjust to make the defendant bear the costs of the proceedings which were forced upon him, and which the claimant is unable or unwilling to carry through to judgment: Brookes v HSBC Plc: Jemitus v Bank of Scotland Plc … [2011] EWCA Civ 354.”

CPR rule 44.2 gives the court a wide discretion as to what costs order should be made and the creative orders that could be made. Rule 44.2 provides,

“(4)

In deciding what order (if any) to make about costs, the court will have regard to all the circumstances, including - ”

(a)

the conduct of all the parties;

(b)

whether a party has succeeded on part of its case, even if that party has not been wholly successful; and

(c)

any admissible offer to settle made by a party which is drawn to the court’s attention, and which is not an offer to which costs consequences under Part 36 apply.

(5)

The conduct of the parties includes—

(a)

conduct before, as well as during, the proceedings and in particular the extent to which the parties followed the Practice Direction—Pre-Action Conduct or any relevant pre-action protocol;

(b)

whether it was reasonable for a party to raise, pursue or contest a particular allegation or issue;

(c)

the manner in which a party has pursued or defended its case or a particular allegation or issue;

(d)

whether a claimant who has succeeded in the claim, in whole or in part, exaggerated its claim; and

(e)

whether a party failed to comply with an order for alternative dispute resolution, or unreasonably failed to engage in alternative dispute resolution.”

Returning to the question of discontinuance and whether the presumption should be departed from. HHJ Waksman QC reviewed the authorities and derived eight principles, which were approved on the appeal from his decision in Brookes v HSBC Plc [2011] EWCA Civ 354. At paragraph 6 Moore-Bick LJ restated those principles, but reduced them down to six, as follows,

“i)

when a claimant discontinues the proceedings, there is a presumption by reason of CPR 38.6 that the defendant should recover his costs; the burden is on the claimant to show a good reason for departing from that position;

ii)

the fact that the claimant would or might well have succeeded at trial is not itself a sufficient reason for doing so;

iii)

however, if it is plain that the claim would have failed, that is an additional factor in favour of applying the presumption;

iv)

the mere fact that the claimant's decision to discontinue may have been motivated by practical, pragmatic or financial reasons as opposed to a lack of confidence in the merits of the case will not suffice to displace the presumption;

v)

if the claimant is to succeed in displacing the presumption he will usually need to show a change of circumstances to which he has not himself contributed;

vi)

however, no change in circumstances is likely to suffice unless it has been brought about by some form of unreasonable conduct on the part of the defendant which in all the circumstances provides a good reason for departing from the rule.”

At paragraph 10,

“It is clear, therefore, from the terms of the rule itself and from the authorities that a claimant who seeks to persuade the court to depart from the normal position must provide cogent reasons for doing so and is unlikely to satisfy that requirement save in unusual circumstances. The reason was well expressed by Proudman J. in Maini v Maini: a claimant who commences proceedings takes upon himself the risk of the litigation. If he succeeds he can expect to recover his costs, but if he fails or abandons the claim at whatever stage in the process, it is normally unjust to make the defendant bear the costs of proceedings which were forced upon him and which the claimant is unable or unwilling to carry through to judgment. That principle also underlies the decision of this court in Messih v MacMillan Williams. There may be cases in which it can be said that the defendant has brought the litigation on himself, but even that is unlikely to justify a departure from the rule if the claimant discontinues in circumstances which amount to a failure of the claim.”

The claimants agree that the starting point is CPR 38.6(1) and the six principles set out above, albeit they referred the court to Nelson’s Yard Management Company v Eziefula [2013] EWCA Civ 235 which quotes the above at paragraph 14.

The claimants rely in particular on principles five and six as being key to their argument. They also quite properly acknowledge how high a hurdle they have to overcome. They referred the court to paragraphs 30 to 32 in Nelson’s Yard Management where Beatson LJ said,

“30.

The hurdle to displace the default rule in CPR Part 38.6(1) is also a high one. A claimant who discontinues must (see Moore-Bick LJ’s sixth principle in Brookes v HSBC Bank) generally show some form of unreasonable conduct on the part of the defendant which provides a good reason for departing from the rule. The height of this hurdle can be illustrated by two examples.

31.

First, the mere fact that a claimant has got all or almost all he could reasonably hope to achieve from the proceedings has been said not to justify a claimant from relying on the avoidance of a trial which would be solely about liability to recover costs as justifying a departure from the default rule: see Patten LJ in Messih v MacMillan Williams [2010] EWCA Civ 844 at [28], [30] and [31]. In Brookes v HSBC Bank, after referring to what Patten LJ stated, Moore-Bick LJ observed (at [10]) that a claimant who seeks to persuade the court to depart from the default rule must provide cogent reasons and is unlikely to be able to satisfy the court that there is good reason to do so save in unusual circumstances. In Messih’s case the achievement by the claimant of what he had sought from the proceedings by a settlement with one of a number of defendants did not justify disapplying the rule when he discontinued against other defendants.

32.

The second example concerns the position of disputed material. It is clear that once there is to be no trial, it is not the function of the court considering costs to decide whether or not the claim would have succeeded: see Re Walker Wingsail Systems PLC [2006] 1 WLR 2194, per Chadwick LJ at [12], and HHJ Waksman’s second principle in Teasdale v HSBC Bank PLC, [2010] EWHC 612 (QB) at [7(2)]. But it is also clear (see Moore-Bick LJ’s sixth principle in Brookes v HSBC Bank)) that it is the function of the court to consider whether the unreasonableness of a defendant’s conduct provides a good reason for departing from the default rule.”

The facts in Nelson’s Yard Management Company v Eziefula provide a very useful illustration of when the court will disapply the usual rule. This was a case concerning an alleged breach of the Party Wall etc Act 1996. The defendant had begun excavating within one metre of the rear wall of the claimants’ property and had failed to serve the requisite Party Wall Act notices. To compound matters the defendant failed to respond to any pre-action correspondence, which had sought access for the claimants’ surveyor and had set out that the works were in breach of the Act. So, the claimants issued proceedings seeking an injunction to stop the works until an award had been agreed or the claimants’ surveyor given access to inspect the foundations. The defendant eventually permitted access to the claimants’ surveyor and also served a party structure notice. A modest award was issued to the claimants for damage to the foundations. The claimants discontinued the claim having achieved what they wished to. The Court of Appeal considered that the claimants were entitled to protect their position and issue the claim and pursue the claim until they ascertained the defendant’s position. The position changed when the defence was filed. At that stage the claimants did not apply to strike out the claim or seek summary judgment on the basis that the defence was spurious. That decision was probably taken because they were unlikely to succeed given the issues in dispute. The court ordered that the defendant should pay the claimants’ costs to the date when the defence was served and thereafter it should be no order as to costs.

The claimants also referred me to a paragraph further on in Beatson LJ’s judgment,

“34.

What is the position where the conduct relied on by the claimant as a ground for departing from the default rule is unrelated to the merits of the claim or the defence? Such a scenario may appear far-fetched. But, leaving that aside, to require the claimant to proceed to a trial where the real issue concerns conduct of the defendant unrelated to the pleaded cases seems a waste of scarce court resources and the resources of the parties.”

This would suggest a link between the unreasonable conduct and the reason for discontinuance, outwith the merits of the claim itself. The claimants also sought to argue that the emergence of new evidence can justify a departure from the usual rule. In Harrap v Brighton & Sussex University Hospitals NHS Trust [2018] EWHC 1063 the defendant’s consultant cardiologist gave evidence in cross examination that did not feature in his witness statement. The evidence was such that it led to the claimant discontinuing, his case on causation was then bound to fail. The Judge varied the usual default costs rule on discontinuance and provided that after the date for exchange of witness statements each party would bear their own costs. This is an extreme example and one can see why the Judge would be satisfied that it was appropriate to disapply the presumption in CPR 38.6.

As to the basis of assessment, under CPR 44.3(1),

“(1)

Where the court is to assess the amount of costs (whether by summary or detailed assessment) it will assess those costs—

(a)

on the standard basis; or

(b)

on the indemnity basis,

but the court will not in either case allow costs which have been unreasonably incurred or are unreasonable in amount.”

Where no basis is indicated in an order the costs will be assessed on the standard basis.

In Excelsior Commercial and Industrial Holdings Ltd v Salisbury Hamer Aspden & Johnson [2002] EWCA Civ 879 the court declined to give detailed guidance on when the court should award indemnity basis costs. However, Woolf LJ and Waller LJ both said that the making of such an order would be appropriate in circumstances where the conduct of the parties or other particular circumstances of the case, or both, took the situation “out of the norm”.

Weakness of the merits of a case is not, without more, justification for costs to be ordered on an indemnity basis. In Arcadia Group Brands Ltd v Visa Inc [2015] EWCA Civ 883 (Footnote: ¹) Sir Terence Etherton considered that the position might be different if the proceedings or steps taken within them were based not only on a plainly hopeless case but were also motivated by some ulterior commercial or personal purposes or otherwise for purely tactical reasons unconnected with any real belief in the claim’s merit.

Where the application for indemnity costs is based on litigation conduct Sir Anthony Colman in National Westminster Bank plc v Rabobank Nederland [2007] EWHC 1742 (Comm) said at paragraph 28,

“Where one is dealing with the losing party’s conduct, the minimum nature of that conduct required to engage the court’s discretion would seem, except in very rare cases, to be a significant level of unreasonableness or otherwise inappropriate conduct in its widest sense in relation to that party’s pre-litigation dealings with the winning party or in relation to the commencement or conduct of the litigation itself. …

But in each case in which the costs of the whole litigation are under consideration, the conduct adversely criticised must be looked at in the context of the entire litigation and a view taken as to whether the level of unreasonableness or inappropriateness is in all the circumstances high enough to engage such an order.”

He also went on at paragraph 29 to refer to the observations of Christopher Clarke J in Balmoral Group Ltd v Borealis (UK) Ltd [2006] EWHC 2531 (Comm),

“The discretion is a wide one to be determined in the light of all the circumstances of the case. To award costs against an unsuccessful party on an indemnity scale is a departure from the norm. There must, therefore, be something – whether it be the conduct of the claimant or the circumstances of the case – which takes the case outside the norm. It is not necessary that the claimant should be guilty of dishonesty or moral blame. Unreasonableness in the conduct of the proceedings and the raising of particular allegation, or in the manner of raising them may suffice. So may the pursuit of a speculative claim involving a high risk of failure or the making of allegations of dishonesty that turn out to be misconceived, or the conduct of an extensive publicity campaign designed to drive the other party to settlement. The making of a grossly exaggerated claim may also be a ground for indemnity costs.”

The nature of the applications and the test that each side must meet on their respective applications is such that I will need to go through the background and the litigation chronology in some detail.

Luis Martin Lazaro (“Mr Martin”) and Daniel de Carvajal Aguilar (“Mr de Carvajal”) are said to have had the idea for software to identify and procure significant flight booking savings, which would be later be developed into Airline Reservation Price Optimisation (“ARPO”) in 2011. It is said to be a unique product comprised of source code including ‘bots’ which can identify and procure significant flight booking savings through an automated process. The claimants’ case was that no third-party product offered the same features as ARPO.

They arranged for the development and commercialisation of the software to be carried out by two companies: from April 2013 Trappit SA, the first claimant, a company incorporated in Panama where the developers were then living; and from June 2014 by Trappit SA and Trappit Tec, the second claimant, a company incorporated in Spain. The third claimant is a Spanish company which employed the developers of the code.

The key features of ARPO are more particularly set out in schedule 2 to the particulars of claim. In short it was an automated process which monitored whether a plane ticket, which had been booked, had subsequently reduced in price and, if so, the ticket would be modified or cancelled and the ticket automatically rebooked at the lower price. Although the rebooking feature does not appear to have been present in the original version of ARPO, or of that version demonstrated to Amex, which lies at the heart of this claim. (Footnote: ²)

The claimants assert that copyright subsists in ARPO and the second claimant is the owner of that copyright. They also pleaded a case that the first and/or second claimant were the owners of confidential information about ARPO which included details of the architecture and set up of ARPO together with the computer code for ARPO.

In late 2013 the first claimant was introduced to Amex Europe, a group of companies (“Amex Europe”), entities of which operated through the brand name American Express Global Travel division. They were interested in potentially selling or licensing ARPO to Amex Europe’s business travel division. The first meeting took place on 3 February 2014.

On 22 February 2014 Mr de Carvajal emailed a general manager at American Express Global Business Travel, Ms Cliffe, asking that a non-disclosure agreement (“NDA”) be signed, and attached a draft which stated that it was governed by Spanish law. Ms Cliffe provided an Amex Europe standard draft instead. The meeting went ahead on 3 March 2014 without an NDA in place.

An NDA was signed on 18 March 2014, using the Amex Europe template but incorporating certain amendments from the first claimant, including that the agreement was to be governed by Spanish law.

The claimants contend that at a meeting on 7 May 2014 a USB memory stick containing a demonstration version of ARPO was provided to Amex Europe. On the claimants’ case this was a “highly confidential source code” “shared with Amex for the purpose of demonstrating how ARPO worked and to enable it to be tested using Amex’s own data” (Footnote: ³). Although the existence and contents of the memory stick were disputed by the defendant. The defendant’s case was that it was only provided with access to an online demo portal, which it did not access, and which provided no access to the source code. (Footnote: ⁴) It is said by the claimants that Amex Europe described ARPO as a “game changer”.

At this time, Amex Europe, which was part of the international American Express Group, operated travel management services under the American Express Global Business Travel Brand. GBT Travel Services UK Limited (“GBT UK”) was created in 2014 as a joint venture between the American Express Group and private equity investors. GBT UK acquired Amex Europe’s travel management services business in the United Kingdom from 30 June 2014.

On 16 July 2014 Ms Verrillo of GBT UK emailed Mr de Carvajal to say that GBT UK was not interested in licensing ARPO.

The claimants’ position is that by assignment on 6 October 2014 all intellectual property and associated rights and all rights of suit were assigned to the first claimant, although it is also their case that prior to this the first claimant had implied rights. A further assignment was entered into on 8 March 2015 between the first and second claimants, although it was not pleaded that it included all of the extensive intellectual property rights that the earlier assignment had, this appears to be more by omission and nothing turns on this for the purposes of the applications.

In February 2015 the claimants received a formal press release (Footnote: ⁵) from Global Business Travel Spain SL, formerly known as American Express Barceló Viajes SL (“GBT Spain”), which advertised the travel savings software called LastFare. This was a product developed by GBT Spain, but not commercialised in the United Kingdom. The significance of this being that copyright is territorial. It appeared to the claimants from the press release that its functions were the same as ARPO. Therefore, the claimants inferred that Amex Europe or an affiliated company had used all or part of the source code for ARPO to develop LastFare.

On 3 June 2015 (Footnote: ⁶) the claimants issued criminal proceedings in Spain against 9 individual defendants and 2 corporate defendants, “American Express España” (Footnote: ⁷) and “American Express internationally”.

In June 2015, after the criminal proceedings had begun, GBT Spain stopped marketing LastFare.

In around March 2017 Amex Europe began to market a product called Air Re-Shop Expert (“ARE”), which also had materially the same features as ARPO. Again, the claimants inferred that Amex or an affiliated company had used all or part of the source code for ARPO to develop ARE.

The Spanish court ordered a computer forensic expert to investigate and draw conclusions on two key matters: (i) assessing whether LastFare plagiarised ARPO; and (ii) comparing their source codes. Mr Francisco Javier Folgado de la Rosa prepared a report dated 22 October 2019. At page 5 he describes the main purpose of the report as separately assessing each software application by drafting an expert review about its characteristics (programming language, hardware infrastructure, etc), functionality and outputs. At paragraph 5.3 he describes the software applications as operating in a different manner with different programming language and different architecture. His findings are that LastFare lacks the capabilities needed to enable GBT Spain to be using it as a best fares search engine, so that LastFare would be able to cover, at most, 10% of ARPO’s capacity. ARPO is an automated end to end process. In contrast LastFare has manual processes. The expert concluded that the source code that he received from GBT Spain was so inferior that it was not created by copying ARPO.

On 15 January 2020 the defendants to the criminal proceedings requested that there be a dismissal with prejudice of the proceedings in view of the investigations. The complainant objected. In a report dated 4 February 2020 the public prosecutor’s office requested the dismissal without prejudice of the case: it being considered that there was not enough evidence to suggest that a crime had been committed.

In Mr de Carvajal’s first witness statement dated 2 December 2020 he referred to evidence being given by Mr Folgado orally in the Spanish court. I understand from his witness statement that he was not present in court but listened to the recording subsequently. He refers to the expert saying that it was approximately 95% probable that GBT Spain was using the source code for ARPO and that he could not be certain because GBT Spain had failed to provide the source code that it was in fact using to him. Although, of course, Mr Folgado was a court appointed expert and his written report and any oral evidence was before the Spanish Court. As was the oral evidence of Mr Dupuy, a witness from GBT Spain who according to the Claimants’ solicitors stated that ARPO and LastFare “fulfil the same function” (Footnote: ⁸).

On 9 April 2020 the claim in England was issued: Mr McElroy describes the issue as being on a protective basis following advice on limitation. This was prior to the conditional dismissal in the Spanish proceedings but after the expert’s report had been received. The claim concerns alleged copyright infringement and misuse of confidential information relating to the ARPO software developed by the claimants. It was the claimants’ case that LastFare and/or ARE were produced by copying from ARPO’s source code and misusing confidential information. In addition, there was an unlawful means conspiracy claim. The prayer sought various heads of relief: damages including damages pursuant to section 97 of the Copyright, Designs and Patents Act 1988 and/or regulation 17 of The Trade Secrets (Enforcement, etc) Regulations SI 597/2018; equitable compensation and/or an account and all necessary inquiries; delivery up of all confidential information; an injunction to restrain any misuse or dissemination of the confidential information; an injunction restraining the defendant from using and marketing its own infringing software; and a declaration that copyright vested in the first and/or second claimant subsists in its software and that the defendant has infringed that copyright.

As is quite common in these claims, paragraph 28 of the particulars of claim pleaded a case based on inference that LastFare was produced by copying a substantial part of the code for ARPO and/or using the confidential information. Although it is relevant to note that breach of confidence proceedings can be used in an oppressive way to harass competitors and ex-employees. A claimant in such a claim must give full and proper particulars of all confidential information on which they intend to rely in the claim. This equally applies to the identification of the material said to be confidential, and that it is indeed confidential. It is the defendant’s case that much of the purported confidential information was already in the public domain, including as a result of the claimants’ published patents and widespread market use over a number of years.

Criticism has also been made by the defendant of the copyright claim. In particular that the works alleged to have been infringed were not sufficiently identified, no specific acts of infringement were alleged, the claimants failed to identify an example of “literal” copying and that there were no particulars of the points of similarity.

It is said that because of the limitation issues the claimants did not have enough time to engage in pre-action conduct correspondence with the defendant. They did attempt post issue to comply with the spirit of pre-action conduct by setting out their claim in full and proposing an extension to the timetable of 56 days for filing the defence to enable the parties to exchange information about the claim and explore the possibility of non-court based dispute resolution. That is contained in a letter dated 5 August 2020 which required a substantive response and extensive disclosure by 28 August 2020. The letter specifically refers to the Spanish criminal proceedings alleging that some or all of the defendants had engaged in criminal fraud, intellectual property crime, unfair competition and crimes related to markets and consumers. Reference is made to a precautionary claim in the Spanish criminal complaint in the sum of €14,681,239. The claimants go on to describe the loss in respect of the English proceedings as either the fair value of the licensing of ARPO which it is said in April 2014 would have had a value of approximately €174 million to €275 million. Alternatively, it is asserted that the defendants are liable to account for the profits made from the misuse of confidential information and/ or copyright information, which is estimated to be between approximately €780 million and €1,210 million. The claimants requested a copy of the source code both for LastFare and ARE.

This letter was sent on 5 August 2020, the same day that the claim form and particulars of claim were served; the period for service of the claim was due to expire on 9 August 2020. Delay is explained, the claimants suggest, by them having to secure after the event insurance in the market, obtained around 3 August 2020. The claimants anticipated that they would be likely to face an application for security for costs.

The defendant responded by letter dated 12 August 2020 complaining that the approach proposed by the claimants was not reasonable and that in the circumstances there should be an immediate stay of 6 months, to enable the parties to investigate the matters referred to in the 5 August 2020 letter and to allow for proper exchanges to take place. In light of the claim and the Spanish criminal proceedings, which were still ongoing at this stage, there is some force in the defendant’s position. Although with the benefit of hindsight I wonder if by then the parties had already become entrenched.

In the hearing the claimants criticised the defendant for failing to deal with confidentiality undertakings that they had sought in a letter dated 30 July 2020, prior to the proceedings being issued. Although I note in the defendants’ letter dated 12 August 2020 they repeated a position they had adopted in a letter dated 3 August 2020 saying that the client would consider confidentiality undertakings if the claimants could explain why it considered them necessary. It was also stated that the claimants should identify with precision the information which they seek to protect and to provide reasonable draft undertakings for consideration by the defendants. Both parties were clearly aware of the significance and need for confidentiality undertakings.

On 14 August 2020 the Spanish Court of First Instance issued its judgment. At page 3 the Judge refers to Mr Folgado’s written report and also his oral evidence. The Judge says that Mr Folgado, “stated that the two programs ARPO/LASTFARE were not comparable, had nothing to do which each other (graphically pointing out that it was like comparing a vehicle to a skateboard )”. Mr Folgado highlighted that ARPO does have an automatic search system which LastFare does not.

The claimants emphasise before this court, as no doubt they did before the Spanish Court, that they were not provided with the source code as provided to Mr Folgado and could not independently verify his evidence.

The Judge noted at this late stage in the proceedings the complainant was now suggesting that another program, ARE, had plagiarised ARPO and that LastFare was a decoy. However, the complainant had submitted an expert report which concluded that LastFare was a copy of ARPO, and that they were completely symmetric products.

“In other words, it now appears to redirect the subject matter of the proceedings to another program, different to that which gave rise to the criminal claim and therefore different to LastFare. … the criminal complainant’s intention consisting in a modification of the subject matter of the proceedings cannot be admitted. ” (Footnote: ⁹)

The court upheld the request for a dismissal without prejudice. The significance of this is that in deciding whether to progress from the investigation phase to the intermediate phase the Spanish Judge is merely considering whether a crime could have been committed, which is a lower threshold than the usual criminal standard in England where a complainant would have to show that it is more likely than not that a criminal offence was committed (Footnote: ¹⁰).

The effect of this provisional dismissal was that the Spanish proceedings were closed but could be re-opened by the parties, the court or the Public Prosecutor, if new evidence came to light which justified it.

The English claim was originally brought against American Express Europe LLC (“Amex Europe”), a Delaware corporation with a registered branch in England and GBT Travel Services UK Limited, an English company.

The then defendants acknowledged service on 19 August 2020 indicating that they would contest jurisdiction.

On 30 September 2020 the defendants applied to strike out or stay the proceedings on the basis that the claimants were contractually bound to litigate the claims in Spain rather than England, or that, in light of proceedings that had already been brought and provisionally determined in Spain, the court should decline jurisdiction or strike out the English proceedings as an abuse of process.

Trappit Tec appealed the provisional dismissal in the Spanish proceedings. On 23 October 2020 the judgment of the Spanish Court of Appeal was issued, dismissing the appeal. They concluded that expert evidence had resolved the criminal case, concluding that the operations of ARPO and LastFare differ and they are different programs. They go on to say at page 6,

“Considering the foregoing, it is not acceptable that, now, the complainant states that this information is irrelevant and attempts to modify the object of the investigation, alleging the existence of alleged crimes of discovery … by the defendants allegedly appropriating the ARPO programme, with the defendants using certain tricks and subterfuges … as said assertions are based on the assumptions and conjectures. There is no indication whatsoever that GBT Spain had access to the ARPO program nor that it was issued with it at any time.”

In Trappit v Amex [2021] EWHC 1344 Snowden J concluded that the jurisdiction clause in the NDA applied to all claims in the English proceedings, although only bound Amex Europe and Trappit SA, as original signatories. Therefore, the court had no jurisdiction over the claims brought by Trappit SA against Amex Europe. The rest of the defendants’ application was dismissed. In particular the Judge did not consider that the essential requirements of an issue estoppel had been met; the decision of the Spanish court was not final and binding so as to prevent future civil claims. At paragraph 173,

“if there is no res judicata in Spain arising out of the provisional dismissal of the Spanish Proceedings, no issue estoppel can arise in England.”

However, he did go on to observe the following points,

“179.

So far as purpose is concerned, the Defendants contend that the English Proceedings claims were commenced after five years of fruitless litigation in Spain, claiming substantial but unparticularised damages, and it can be inferred that they are designed to exert leverage upon American Express to reach some form of settlement contrary to the intrinsic merits or true value of the dispute.

180.

Those are powerful points and well made. Trappit Tec has clearly pursued persons associated with the Defendants extensively in the Spanish Proceedings over a considerable period of time, but appears to have got nowhere in establishing the fundamental allegation of plagiarism of the ARPO program at the heart of its complaints.

181.

The points made by the Defendants have given me substantial pause for thought. So too has the point that, for reasons that I have explained, the NDA prevents the claims by Trappit SA against AmEx Europe from being litigated in England. Although that factor is not relevant to the analysis under Article 25 for the reasons set out in paragraph 70 above, it does point to unified civil proceedings in Spain as being the most appropriate to resolve all the issues between the parties.

182.

As such, I regard the English Proceedings as very close to the margins of an abuse or misuse of the process of the English courts. However, on balance I do not consider that they are sufficiently unfair to the Defendants or risk the administration of public justice in England being brought into disrepute so as to warrant striking them out as an abuse.”

As Mr Justice Snowden did note though, the power to strike out on the basis of repeat proceedings for vexation is conceptually different from the power to strike out a hopeless case. This is a point emphasised by counsel for the claimants, that if the claim has been so hopeless from the outset, as maintained by the defendant, then the defendant would have applied to strike out or for reverse summary judgment. I also note that the claimants in their submissions accept that there was a factual dispute about what was provided to Amex Europe on 7 May 2020 (Footnote: ¹¹).

It is suggested by the claimants that the Spanish proceedings are irrelevant. I cannot see how that argument can be maintained. They are an important part of the background. Whilst it is true that the defendants failed to strike out the English claim on the basis of issue estoppel, that of itself does not mean that the evidence given in the Spanish proceedings can simply be disregarded, and that it would not have formed part of the evaluative process of deciding whether the English proceedings should be issued, served and pursued.

On 27 August 2021 the defendant served a Part 18 request for further information, which extended to 101 requests. On 13 September 2021 the claimants again suggested exchanging the source code. The defendant agreed in a letter dated 16 September 2021, but said it was premature before the defence was served.

On 22 October 2021 the claimants provided responses to the Part 18 request, or those that they considered were necessary in order for the defendant to formulate its defence. The defendant considered that the pleading in respect of Confidential Information and ARPO remained inadequately particularised, although no application to strike out the claim on this ground was made.

On 19 November 2021 the claimants served re-amended particulars of claim.

On 17 December 2021 the defendant served its defence. In summary its case in respect of LastFare was that it was independently created by GBT Spain and was not copied or derived from ARPO. In respect of ARE, this was said to have been commissioned from Yapta Inc, based on an existing Yapta product known as FarelQ, and a competitor of the claimants. At paragraph 2.2 of the original defence it is pleaded,

“ARE was not copied or derived from ARPO but rather from FarelQ. GBT UK did not write, nor does it have or has it ever had access to or the right to obtain, the source code for ARE.”

The Yapta licence was annexed to the defence, albeit in redacted form.

On 4 February 2022 the claimants’ solicitors wrote about the practical aspects of exchanging source code and on 21 February 2022 proposed simultaneous exchange of the source code for ARPO and LastFare. On 21 February 2022 the claimants also served their reply and a Part 18 request for information.

In a letter dated 4 March 2022 the defendant’s position remained that they did not object to exchange but asked what the claimants’ proposals were in respect of ARE, particularly in light of the claimants’ reply which it was said did not meet head on the defendant’s case in respect of ARE. The defendant also stated that it had no legal access to the ARE source code. At this stage and notwithstanding the defence the claimants were maintaining a claim in respect of ARE.

The correspondence appears quite circular on these issues. Again, in correspondence on 7 March 2020 the claimants’ solicitors sought the exchange of the ARPO and LastFare source code. The defendants reiterated its position in a letter dated 16 March 2022.

On 28 March 2022 the defendant served its replies to the Part 18 request.

On 10 May 2022 the claimants’ solicitors confirmed that the ARE claim would no longer be pursued. They again sought to exchange the source codes of LastFare and ARPO. After chasing on 24 May 2022, the defendant’s solicitors replied on 31 May 2022,

“GBT remains willing, in principle, to exchange this code with your clients, however before it commits to the costs associated with commissioning experts to review source code, please confirm what further purpose the Claimants consider that the exchange of LastFare and ARPO source code would serve at this stage, in particular given Mr Folgado’s clear conclusion that LastFare was not plagiarised from ARPO.”

It is said on behalf of the claimants that there was considerable delay by the defendant in providing an unredacted copy of the licence agreement in respect of ARE. These redactions were removed on 21 January 2022 (Footnote: ¹²).

An order was made by consent on 29 June 2022 by Deputy Master Raeburn permitting the amendments that removed the ARE claim and ordering the claimants to pay the defendant’s costs of the discontinued claim on the standard basis. I do note that, despite criticism by Mr McElroy in his witness statements, the claimants did not seek to disapply the usual rule, and conversely the defendant did not seek indemnity costs in respect of the ARE claim. I also note that there was a delay between the unredacted licence agreement being provided, although I make no comment about whether this was necessary, and the discontinuance of the ARE claim.

Thereafter the claim continued in respect of LastFare only. By this stage the Spanish criminal proceedings had concluded nearly 2 years previously, and those related explicitly to the allegations concerning LastFare. Although the claimants maintain that they only knew that the claim would fail shortly before the notice of discontinuance was served, when “they were reasonably satisfied that their computer programming expert had been provided with all of the relevant source code for LastFare and he had concluded that it had not been copied from ARPO.” (Footnote: ¹³)

Going back to the litigation chronology, on 30 June 2022 the claimants served re-re-amended particulars of claim removing the ARE claim. On 14 July 2022 the defendant served its amended defence.

A costs and case management conference had been listed on 20 and 21 July 2022. The parties’ solicitors continued with their circular correspondence about the LastFare source code. It is suggested by the claimants’ solicitors that the defendant was effectively finding another excuse not to disclose the source code, by now raising an issue about security for costs, which they had vacillated about pursuing. Although the defendant made the application on 31 May 2022. Two observations follow from that. First the claimants had themselves been aware of the risk of an application for security for costs at the outset, which was the reason why they say they delayed in serving the claim to obtain after the event insurance. Second, on the claimants’ case the defendant was coming up with excuses not to disclose the source code for LastFare, but I ask rhetorically to what end?

On 20 July 2022 Deputy Master Arkush gave case management directions which provided for a split trial on liability and extended disclosure to be given by 19 November 2022. Issue 15 in the Disclosure Review Document was defined as, “Between 1 February 2014 and 28 February 2015, what electronic files made up LastFare?” Model C extended disclosure was ordered. In correspondence prior to the hearing the claimants criticised the defendant for changing the parameters of disclosure of the LastFare source code, saying that they had previously sought disclosure of all source codes preserved by GBT Spain and now the defendant was only offering to disclose the version of the source code which GBT Spain had provided to the defendant. Although I note that the claimants agreed to this, albeit it is said with reluctance. At this stage the claimants’ incurred costs were £2.5 million and budgeted costs of £9.8 million. The court approved costs including incurred and estimated costs totalled £7.3 million. The defendant’s incurred costs were £1.06 million and budgeted costs £4.09 million (Footnote: ¹⁴). The court approved costs including incurred and estimated costs totalled £4.78 million.

On 21 July 2022 Deputy Master Arkush ordered the claimants to provide security for costs by means of an after the event insurance policy and anti-avoidance endorsement cover in the form of staged cover (Footnote: ¹⁵); the principle but not the quantum having been agreed between the parties.

On 8 August 2022 the defendant provided the source code for LastFare, as agreed. This was either 4 or 7 days after the claimants implemented the first tranche of the security for costs. The claimants’ solicitors questioned in a letter dated 1 September 2022 whether the source code related to LastFare because the functionality did not match the press release. Although the functionality of LastFare reflected Mr Folgado’s expert analysis in the Spanish proceedings.

On 16 September 2022 the claimants applied for a disclosure guidance hearing, to be determined without a hearing. Deputy Master Rhys considered it was premature to have such a hearing, particularly in light of the amount of material placed before the court, some 80 pages, and that extended disclosure was not due until 19 November 2022. On 16 November 2022 the parties agreed keyword searches.

There was then further correspondence between the parties culminating in the defendant obtaining an extension to the extended disclosure deadline to 27 January 2023; the order of Deputy Master Smith dated 5 December 2022.

On 27 January 2023 the parties provided extended disclosure, which included the defendant providing the further source code for LastFare to the claimants’ experts. The defendant also disclosed a PowerPoint presentation created by GBT Spain on 16 February 2015, which compared ARPO’s performance with LastFare’s in relation to bookings for one of GBT Spain’s customers. That presentation suggested that LastFare had greater functionality than ARPO. By this stage the defendant says that it had provided additional LastFare code, dated 17 December 2014, 21 January 2015 and 25 February 2015. In addition, and outside the parameters of issue 15 the defendant provided versions dated 28 May 2015, 27 November 2015 and 24 February 2016.

By letter dated 24 March 2023 the claimants’ solicitors questioned whether disclosure was complete because it was said that the functionality of the code still did not match the previous claims in the press release and now in the PowerPoint presentation as well. On 17 March 2023 the defendant requested a statement of case on infringement and set out at some length in what ways it considered the claimants’ pleadings to be deficient.

On 24 March 2023 the claimants raised technical questions about the disclosed source code. On 21 April 2023 the defendant’s solicitors replied stating that LastFare required manual intervention in order to achieve the reported results. The claimants have emphasised that this is the first time in the English proceedings that this was acknowledged by the defendant. Although of course that position had been clearly set out in Mr Folgado’s report and accepted by the Spanish Court, at first instance and on appeal. Also, on 21 April 2023 the defendant’s solicitors provided the current version of the LastFare source code. The defendant’s solicitors also reiterated that the defendant had no involvement in the development or operation of the LastFare tool, which had now been exhaustively dealt with in both the Spanish and English proceedings.

The claimants remained suspicious about the defendant’s disclosure and position. Mr McElroy accepts in his evidence that manual intervention could explain the contentions about functionality in the PowerPoint presentation but not in the processing of some 90,000 tickets in November 2015. This assertion was put in the claimants’ solicitors letter dated 10 May 2023. Although by this stage the claimants had the source code disclosed in January 2023 and further source code disclosed in April 2023. The defendant’s solicitors did not respond to this letter.

On 19 May 2023 the defendant served its application requiring the claimants to provide particulars of infringement.

The claimants’ solicitors assert that the claimants were only in a position to know that their claim could not succeed when their expert reported to them on 16 June 2023. The notice of discontinuance was then filed on 29 June 2023. Their case is that the source codes disclosed in January 2023 and April 2023, together with the defendant’s solicitors letter dated 23 April 2023, enabled the claimants and their experts to take an informed view on the prospects of the claim. Although the contents of the defendant’s solicitors’ letter did not provide new information that the claimants were unaware of before. Mr Howarth in evidence also suggests that even on the claimants’ case the source code disclosed in January 2023 would have clearly shown that the claim was doomed to fail.

Needing to go through the litigation chronology in this detail demonstrates why it should be such a high hurdle for the claimant to establish that the default position on discontinuance should not apply and equally why seeking indemnity costs requires the case to be “out of the norm”. Steps taken in the proceedings may be relevant at the time but taken out of context may be given greater significance. Furthermore, like the owl of Minerva it is easier to understand matters with the coming of dusk. This hearing is not a trial nor an application to strike out the claim on the basis that it discloses no reasonable grounds for bringing the claim nor a reverse summary judgment application showing that the claimants had no real prospect of succeeding on the claim. Indeed, I note that in Brookes v HSBC the third principle is described as an additional factor, not determinative of the application. What I have set out at some length is commercial litigation, hard fought, with deep mistrust.

The claimants’ case on why the presumption under CPR 38.6 should be dis-applied is advanced on 3 primary bases, although these interlink and are founded on the timing of the disclosure of LastFare’s source code.

I bear in mind that the claimants’ counsel accepts that the claimants must show some form of unreasonable conduct on the part of the defendant which, in the circumstances, provides a good reason for departing from the usual rule.

Firstly, it is said that it was reasonable for the claimants to pursue the claim until they had received all of the LastFare source code. The claimants’ solicitors maintain that was on 21 April 2023 together with the contents of the defendant’s solicitors letter dated 21 April 2023.

Secondly, acting it is said reasonably the defendant would have disclosed the LastFare source code as early as possible in the proceedings.

Thirdly, had the defendant disclosed the LastFare source code earlier the claimants would have realised earlier that their claim was bound to fail and would therefore have discontinued earlier.

One of the points that the defendant has sought to drive home in this application is that the claim was doomed from the outset. By the nature of these types of claims they are often based on inference, the relevant information being in the control of the defendant or as in this case a related entity of the defendant based in another jurisdiction and, before discontinuance of the ARE part of the claim, a third party. They can be difficult cases to run successfully. So, saying that it does not appear that the claimants had a strong claim, albeit they had a strong belief that the defendant had behaved in such a way that the claim was meritorious to bring, both in respect of the criminal proceedings in Spain and in relation to the English claim. The defendant also points to the fact that Mr Justice Snowden had indicated that the claim was on the borders of being abusive, that perhaps the better jurisdiction was to pursue this in Spain. However, both in the Spanish proceedings and in the judgment of Mr Justice Snowden there was an acknowledgement, certainly in respect of the defendant, that the claim could be litigated in this jurisdiction. The claimants had obtained after the event insurance prior to service of the claim. That would have involved an evaluation of the merits of the case by the underwriter and that would usually require a report or opinion from a lawyer outlining the strengths and weaknesses of the case.

As I have already indicated I do not accept that the Spanish proceedings can be ring fenced and put to one side as being irrelevant, they are part of the essential factual matrix. Although the direction and control of that inquisitorial process was firmly in the hands of the Spanish court. The claimants were not given access to the source code for LastFare in those proceedings, although a court appointed expert did have access. The claimants would not accept that he had access to the correct version of the source code, which in their belief would have established the civil wrongdoing on the part of the defendant.

It is pertinent though that the defendant did not bring an application to strike out or for reverse summary judgement on the basis that the pleadings disclosed no reasonable grounds for bringing the claim or did not disclose a real prospect of success and there was no other compelling reason for the matter to proceed to trial. I do not criticise the defendant with the benefit of hindsight and an ability to survey the entire landscape of this litigation from analysing now the merits of the claim but it is not determinative of this application. I do accept that there were weaknesses in the claim, albeit it was predicated on the basis of a press release that certainly suggested remarkable similarities in functionality between ARPO, which the claimants say was novel in the marketplace although this is disputed by the defendant, and LastFare.

The defendant certainly had an interim application in mind because they had provided for it in the costs budget. I can see why though the defendant considered that there would be a risk in pursuing an interim application and elected not to bring such an application. There was a range of issues between the parties and, for example, a central factual dispute as to what was provided to Amex Europe at the meeting on 7 May 2014. There were disputes such as whether the claimants’ “confidential information” was in the public domain and therefore not confidential at all, putting to one side the points that the defendant raised about the lack of specificity in the pleadings. Whilst I do not accept the claimants’ argument that I am to infer from the fact that the defendant did not bring an application that the claim had a reasonable prospect of success I do consider it relevant that the defendant elected not to bring an application.

So really the claimants’ application turns on whether the failure by the defendant to provide early disclosure of the source code is a good reason to dis-apply the usual costs rule. I have already set out in detail the relevant factual background and the litigation history. To suggest that the entirety of the LastFare part of the claim turned on the provision of the source code is to take too simplistic an approach to this heavyweight, highly costly commercial litigation. If that was all that it would take, I question why the claimants did not either engage in pre-action correspondence or accept the suggestion by the defendant in its letter dated 12 August 2020 that there should have been a stay of six months, why they did not seek specific disclosure of the source code at an early stage. The claimants have fallen into the same trap as the defendant in analysing the litigation, which effectively started in 2015 with the Spanish proceedings, with the benefit of hindsight.

This is compounded by the claimants in this application appearing to present a case that they required 100% certainty that the ARPO source code had not been copied and used in the LastFare software before discontinuing the claim. For example, they rely on the fact that the defendant stated in a letter dated 21 April 2023 that LastFare required manual intervention. That was presented as one of the key parts of the decision to discontinue. However, the defence, which was signed with the statement of truth, expressly pleaded that LastFare “relied on manual entry of the tickets data, … a socket-based connection to Amadeus only, and sent notifications requiring manual ticket rebooking” (Footnote: ¹⁶). Mr Folgado in his report to the Spanish court specifically referred to manual processes within the LastFare system. There is also no adequate explanation as to how the LastFare code disclosed in January 2023 was different to that disclosed in April 2023, whether automation features had been present in the earlier version but not in the latter version. Indeed, the defendant had provided the LastFare source code, that it agreed with the claimants to provide and pursuant to the order of the court, in August 2022.

For the reasons that I have set out over the course of this judgment the claimants have not satisfied the high bar that is rightly set for parties in applications such as these, they have not adduced cogent reasons that would justify the court disapplying the usual rule under CPR 38.6. The claimants should pay the defendant’s costs in accordance with CPR 38.6(1).

The matters that I have already dealt with are also directly relevant to the defendant seeking costs assessed on an indemnity basis. For convenience I summarise those as follows: the failure to follow pre-action protocol, that Mr Justice Snowden had considered that the claim was close to the margins of abuse, arguing a case in respect of LastFare that had already failed in Spain, raising the claim in respect of ARE when the claimants could have done so within the Spanish criminal proceedings , the inadequacies of the pleaded claim, the failure to provide security, the failure to provide particulars of the infringement claim and broad requests for disclosure from a third party.

I have already dealt with these within the context of the litigation history above. I have also noted that the defendant elected not to bring an application to strike out the claim for the deficiencies it now relies on so ardently, or to strike out the claim or obtain reverse summary judgment because the claim was so hopeless from the outset, as it now presents. Some of the criticism relates to discrete periods within the litigation, some concern the ARE aspect of the claim which the parties have already agreed would be costs on a standard basis.

I do not consider it feasible or just to analyse this claim by pulling out the LastFare claim in isolation and applying a microscope to points in time. The Spanish proceedings cannot be effectively transposed over the English proceedings, they are entirely different proceedings, one criminal, one civil, conducted in entirely different procedural and evidential ways. Mr Justice Snowden concluded that issue estoppel could not arise in the English proceedings and he allowed the claim to proceed against the defendant. It was obvious that the claimants recognised very early on two issues that would arise, the need for confidentiality undertakings (which applied both ways) and the need for after the event insurance to meet the inevitable application for security for costs.

I am not satisfied that when one steps back and looks at this litigation that it is “out of the norm” that would warrant costs being assessed on an indemnity basis. Costs should therefore be assessed on a standard basis.