Neutral Citation Number: [2020] EWHC 657 (Ch)Case No: HC-2012-000023
IN THE HIGH COURT OF JUSTICE
BUSINESS AND PROPERTY COURTS OF ENGLAND AND WALES
COMPETITION LIST (ChD)
The Rolls BuildingRoyal Courts of JusticeLondon, EC4A 1NL
Before:
MR JUSTICE ROTH
Between:
INFEDERATION LIMITED Claimant - and - (1) GOOGLE LLC
(2) GOOGLE IRELAND LIMITED
(3) GOOGLE UK LIMITED Defendants
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
MS. SARAH FORD QC and MS. CHARLOTTE THOMAS (instructed by Hausfeld & Co LLP) for the Claimant
MR. JON TURNER QC and MS. JULIANNE KERR MORRISON (instructed by Bristows LLP) for the Defendants
Hearing date: 13 November 2019
- - - - - - - - - - - - - - - - - - - - -
Approved Judgment
I direct that pursuant to CPR PD 39A para 6.1 no official shorthand note shall be taken of this Judgment and that copies of this version as handed down may be treated as authentic.
.............................
MR JUSTICE ROTH
Mr Justice Roth:
Introduction
These are substantial competition law proceedings. Three confidentiality rings (or “clubs”) have been established by orders of this court, giving effect to agreement between the parties: a confidentiality ring, which includes the founding members of the claimant (“Foundem”), Mr and Mrs Raff; an inner confidentiality ring, referred to as the “LEO” [legal eyes only] ring, although that term is misleading as the ring comprises not only external solicitors and counsel but also economic experts; and a still more restricted inner confidentiality ring, referred to as the “RLEO” ring, which presently comprises, on the side of Foundem, 10 named external solicitors and counsel.
This is an application by Foundem for admission of an independent expert to the LEO and RLEO rings. Such an application might not normally be controversial. The facts that both sides appeared on this application by leading and junior counsel, and that the extensive argument addressed to the Court took up a full day, show that it is a significant issue which is strongly contested. To understand the reasons and provide a context for the application, it is necessary to describe the nature of the underlying proceedings.
The proceedings
Foundem has since January 2006 operated an on-line business providing a “vertical” search engine that allows consumers to compare prices for goods and services offered on third party websites. Foundem alleges that the defendants (“Google”) have abused a dominant position contrary to Article 102 of the Treaty on the Functioning of the European Union (“Article 102”) and the Chapter II Prohibition in section 18 of the Competition Act 1998. The claim was issued in 2012, but some years before Foundem had complained about Google’s conduct to the European Commission (“the Commission”), which opened an investigation. That investigation was prolonged, but on 27 June 2017 the Commission finally adopted a decision in Case AT.39740 Google Search (Shopping) (“the Decision”) finding that the first defendant, Google LLC (formerly, Google Inc.), had infringed Article 102 by positioning and displaying more favourably in its general search results pages Google’s own comparison shopping service compared to competing comparison shopping services (such as that operated by Foundem). That infringement was found to have commenced, as regards the UK, in January 2008. Google appealed the Decision to the EU General Court and the oral hearing of that appeal has very recently taken place: Case T-612/17 Google and Alphabet v Commission. It is likely to be many months before the General Court delivers its judgment, and it is very possible that there may be a further appeal to the Court of Justice of the European Union.
Foundem has made a number of successive amendments to its Particulars of Claim. On 30 October 2018, Foundem served a draft Re-Re-Amended Particulars of Claim
(“RRAPOC”), including amendments to take account of the Decision. On 22 February 2019, a revised RRAPOC was served with additional references to the recitals in the Decision. In the light of the Decision, the claim which Foundem seeks to advance has become a “hybrid” claim: in part, it is a follow-on claim relying on the infringement of Article 102 found in the Decision; in part, it is a stand-alone claim alleging further grounds of infringement.
The abuse now alleged, or sought to be alleged by the RRAPOC, effectively comprises 10 claims, summarised as follows:
“Search Penalty Without Objective Justification”: Between around 27 June 2006 and 1 December 2009, Google applied Algorithm A to Foundem’s site without objective justification.
“Search Penalty Discrimination”: Google applied Algorithm A without any objective justification to Foundem’s vertical search services, but not to Google’s own vertical search services and/or Foundem’s competitors.
“Other Penalties Without Objective Justification”: Google may have applied further penalties to Foundem’s site from 26 June 2006 onwards.
“Other Penalties Discrimination”: Google applied these other penalties in a manner that discriminated against Foundem’s website as compared to Google’s own vertical search services.
“AdWords Penalty”: Between 1 August 2006 to around 25 September 2007, Google applied a penalty in relation to Foundem’s AdWords account without objective justification.
“AdWords Penalty Discrimination”: Between 1 August 2006 to around 25 September 2007, Google discriminated against Foundem in the Quality Scores given to Foundem’s webpages, as compared to those given to some of Foundem’s competitors.
“Procedural Abuse”: Google failed to respond to Foundem’s requests for algorithmic penalties to be lifted.
“Universal Search Abuse”: Google used its Universal Search to give preferential search rankings to Google’s comparison shopping services over those of its competitors.
“Panda Penalty”: From around April 2011, Google applied its Panda algorithm to Foundem’s website, significantly lowering the position of Foundem’s website (including its comparison shopping services, and its other vertical search services).
“Single Overall Abuse”: further or alternatively, Google’s conduct and its effects as set out above were interrelated and together constitute an ‘overall abuse’.
Foundem contends that its case, with the exception perhaps of the allegation of procedural abuse in claim (g), concerns a series of inter-related ways in which Google applied various algorithmic penalties or discrimination against its website, along with Google’s Universal Search which unjustifiably favoured Google’s own comparison websites. However, the Decision found an abuse only in respect of conduct falling within claims (b), (h) and (i), although Foundem seeks to rely on what is said in the Decision to support some of its other claims.
Foundem therefore accepts that of these claims, (a), (c), (d), (e), (f), (g) and (j) are stand-alone claims. So also is claim (b) insofar as (1) the discrimination is alleged for the period 27 June 2006-December 2007, (2) it concerns alleged discrimination between Foundem’s other (non-shopping) specialised search services and Google’s other specialised search services, and (3) it relates to discrimination between
Foundem and Foundem’s competitors; and claim (i) insofar as it alleges that the
Panda penalty was applied to Foundem’s other vertical search services. In addition, Google contends that there is a stand-alone element to claim (i) to the extent that it implies that application of the Panda algorithm was not objectively justified. It is unnecessary for the purpose of the present application to determine whether that implication arises under claim (i) and constitutes a stand-alone element.
It should be self-evident that most of the claims concern the operation and effect of Google’s complex search algorithms. As Ms Ford QC submitted for Foundem, they will therefore involve evidence of a highly technical nature.
It has been common ground for some time that the present action cannot proceed to trial before the EU proceedings are concluded (unless perhaps any further appeal is restricted to the question of penalty). However, the proceedings in this court are not stayed and various steps may be taken in the meantime. So far, some limited disclosure has taken place.
Google’s pending application
On 17 May 2019, Foundem applied for permission to make the further amendments to its pleading as set out in the RRAPOC. Many of those draft amendments are not opposed, in particular insofar as they plead the contents of the Decision. However, in response, Google issued an application on 18 July 2019 directed at the pleaded standalone claims. Google’s application seeks to strike out, pursuant to CPR rule 3.4(2)(a), those parts of the existing Foundem pleading that advance the stand-alone claims as disclosing no reasonable grounds for bringing those claims; alternatively, it seeks ‘reverse’ summary judgement on those claims. On the same basis, Google opposes Foundem’s application to amend insofar as it relates to the stand-alone claims.
Google also seeks to have the second and third defendants removed as parties, but that is not material to the present application.
Google’s strike out/summary judgment application (“Google’s pending application”) notes that many of the stand-alone claims had also been the subject of Foundem’s complaint to the Commission, but the Commission did not take them forward; and it asserts that insofar as those claims were not raised with the Commission, they are speculative and without foundation. However, it is well established, and Google accepts, that there is no legal impediment under EU law to a claimant in a private action alleging an infringement which goes beyond the scope of an infringement found by the Commission. While Google relies in support of its application on various statements made by the Commission as persuasive, it does not suggest that
they are binding on this court. The fundamental basis of its application is stated to be that “[t]here is no factual and/or legal basis for any of these claims.”
In support of its application, Google has served three witness statements from: (i) Ms Sophie Lawrance of its solicitors; (ii) Mr Cody Kwok, a principal software engineer at Google; and (iii) Mr Michael Pohl, a senior staff software engineer at Google. The evidence of Mr Kwok and Mr Pohl is largely of a technical nature and concerning the aims and operation of the Google algorithms at issue in these proceedings. When served, 20 of the 25 exhibits to Google’s evidence were designated as either confidential, LEO or RLEO. All three witness statements have been designated as confidential.
As I understand it, Google has more recently re-designated the LEO material in its evidence as “confidential”. As a result, Mr and Mrs Raff, who have a good understanding of the relevant technical matters, can see it and so can an expert who is admitted to the (outer) confidentiality ring. But of six documents designated as RLEO, two are exhibited to Mr Kwok’s witness statement and the other four are referred to in his statement. I was taken to those documents in a closed session of the hearing: one document is in fact a less redacted version of another (even with the RLEO restriction, the documents contain some redactions) so in practical terms there are five RLEO documents which are relied on in Google’s evidence.
I was told that of the approximately 1000 documents disclosed by Google, 127 documents are now designated LEO. This follows re-designation of formerly LEO documents to simple confidentiality documents after requests from Foundem.
The date initially fixed for the hearing of Google’s application was vacated because of its appeal to the General Court. It has now to be re-fixed.
Foundem’s current application
Foundem has identified and appointed an independent expert, Mr Philipp Klöckner, and seeks an order that he be admitted to the confidentiality rings. Google has, albeit only on 11 October 2019, agreed that Mr Klöckner can be admitted to the outer confidentiality ring, but refuses to agree to his admission to the LEO and RLEO rings. Foundem wishes their expert to be admitted to the inner rings not only for the immediate purpose of addressing the evidence on Google’s pending application, but more generally for assistance in assessment of the disclosure still to be provided by Google and, eventually, in preparation for trial.
Google submitted that the issue should be considered in the first instance as regards its pending application. If that application succeeds, then the stand-alone claims will fall away, and since the proceedings will then become follow-on proceedings concerned only with causation and quantum, the nature of the disclosure required, and therefore any issues of confidentiality, will be very different.
Moreover, Google further contended that if the issue is confined to its pending application, there is no need to show the RLEO documents, which contain highly sensitive, confidential information, to a technical expert. That submission is advanced by reference to the way that Google puts its case. Its counsel’s skeleton argument states:
“[Google] refers to the RLEO material for two limited purposes only:
i) to ensure that the Court has a sufficient picture of the sort of technical “signals” that [Google] uses to underpin its search engine, and which it is not the task of the competition authorities or courts to re-design in the absence of some departure from competition on the merits;
ii) to show the Court that, by and large, exactly the same evidence and submissions were presented by [Google] to the Commission for the purposes of its own exhaustive multi-year investigation into [Foundem]’s parallel complaint in that forum, following which the Commission decided not to take [Foundem]’s allegations forward. The Commission did not need to instruct any expert or trade witness to reach the conclusion that it is not the province of antitrust to interfere with [Google]’s ordinary product design choices or how its algorithms work, and therefore not to pursue [Foundem]’s allegations in this regard.” And further:
“The key reason why [Google] now refers to this RLEO material at all in support of its application for summary dismissal is simply to show the Court – directly and plainly – that this is evidential material which was considered by the Commission as part of its own independent assessment of
[Foundem]’s claims, and which the Commission did not see as giving rise to any abuse in itself.”
In his oral submissions, Mr Turner QC for Google stressed the limited scope of the Decision, referring to recital (661) which states:
“… the Commission is not preventing Google from applying adjustment mechanisms. (Footnote: 1) The abuse established by this Decision concerns simply the fact that Google does not apply these mechanisms in the same way to Google’s comparison shopping services and competing comparison shopping services.”
Mr Turner explained that Google would essentially be making two points:
That “there is nothing that is, in principle, capable of being abusive in the use of adjustment mechanisms by a search engine. This is competition on the merits.” It is no less competition on the merits if it ends up with a particular website dissatisfied at how it is being ranked.
That if Google had operated a commercial policy to attack rivals, that would be different, but there is no basis for any such suggestion. The case advanced by Foundem is that the measures adopted by Google were not reasonable and proportionate, and went beyond what was necessary “to achieve the aim of dealing with websites that, for example, have low levels of original content.” Such alleged conduct is not capable of constituting an abuse under competition law. There is no “objective benchmark of relevance” according to which Foundem’s website should have a particular ranking.
Mr Turner asserted that the RLEO documents relied on by Mr Kwok did not go to either of those points or to the aim of Google’s algorithms but only, as stated in the skeleton quoted above, to show what material was before the Commission. As for Mr Kwok’s witness statement, Mr Turner stated:
“What he is seeking to do is both show what the Commission received, and he does exhibit that material. As I say, he does not rely on these signals which are protected by the RLEO designations as part of a positive case. But [he] also helps explain to the court, to the judge at the hearing itself, as background, how this all works, because it is an important case and it was essentially considered to be important that the court should also be aware of these basic matters; but as context and not in order to say there is [no] legitimate case being mounted against us which depends on the reasonableness or proportionateness of what is done, and we will be meeting it on the substance.”
Mr Klöckner
If, contrary to Google’s main argument, Foundem would in principle be entitled to have an expert admitted to the LEO/RLEO rings to see the restricted material, Google nonetheless objects specifically to Mr Klöckner on the basis of his business activities and interests. Mr Klöckner is a ‘search engine optimisation’ (“SEO”) consultant based in Berlin, who advises clients on how to improve their rankings in the major search engines. He is said to have more than 14 years’ experience in this field and significant expertise in the field of comparison shopping websites, which is of course the subject of this case.
Google stresses the importance of keeping the signals and techniques which it uses for ranking of search results confidential from the SEO industry, otherwise their value would be lost. Google’s objection to Mr Klöckner is set out in the witness statement from Ms Lawrance of its solicitors, as follows:
“63. Neither I nor Google suggest that Mr Klöckner would deliberately breach any undertaking. The concern is that once information of the type included in the LEO/RLEO documentation is known, it is difficult to see how someone working in the online search and search optimisation industry or advising in the area of online advertising could avoid even unconsciously relying upon it.
64. Thus, granting access to information designated as RLEO would enable those individuals to use their privileged knowledge of the signals used by Google to rank its Search results to secure an advantage that other web designers would not have – and thus it would allow them to ‘game’ and potentially distort the Search results to their advantage. The integrity of Google’s ranking processes relies upon all webmasters or website owners having the same degree of access to information about Google’s ranking (webmasters do have access to the Webmaster Guidelines, and other public material made available to assist them). This will no longer be the case if information of this kind is made available to some individuals offering commercial services to assist companies to improve their Search ranking.
65. To the extent that the LEO information relates to information other than algorithms or signals, it relates to other technical details or to Google’s internal processes and strategic thinking, the disclosure of which raises the same or similar concerns.”
Google raised some additional concerns surrounding Mr Klöckner’s financial interest in a company called Visual Meta, which is a German comparison shopping service that was a complainant in the Commission’s investigation and is itself a subsidiary of comparison shopping company that is bringing a damages claim against Google in Germany. On his LinkedIn page, Mr Klöckner described his role at Visual Meta as including “Advising in the EU competition case” against Google. Google further notes that Mr Klöckner appears to have as one of his clients the website business Trivago, which also submitted a complaint to the Commission against Google.
However, Google was made aware that Mr Klöckner’s interest in Visual Meta is limited to a 4% shareholding, and he clarified by letter dated 11 November 2019 that although he had advised Visual Meta in the past as regards its complaint to the Commission, that work ended in 2016, well before the Decision, and that he is not advising them or anyone else (apart now from Foundem in this action) in any EU competition case against Google. Further, while he includes Trivago in his marketing material among the names of the companies he has advised, that is only because it is a well-known brand (in Germany); in fact, he has not provided Trivago with services since 2015.
Principles
In Al Rawi v Security Services [2011] UKSC 34, the Supreme Court considered whether the court could permit the use of a closed material procedure for a civil claim for damages, whereby material disclosed by the defendant would be seen only by special advocates appointed for the claimants and by the court, but not by the claimants or their own legal advisors. At the outset of his judgment, Lord Dyson JSC set out what he described as some features of a common law trial “which are fundamental to our system of justice (both criminal and civil)”, including that trials are conducted on the basis of natural justice. Lord Dyson stated, at [12], that one aspect of natural justice was this:
“A party has a right to know the case against him and the evidence on which it is based. He is entitled to have the opportunity to respond to any such evidence and to any submissions made by the other side. The other side may not advance contentions or adduce evidence of which he is kept in ignorance. The Privy Council said in the civil case of Kanda v Government of Malaya [1962] AC 322, 337:
If the right to be heard is to be a real right which is worth anything, it must carry with it a right in the accused man to know the case which is made against him. He must know what evidence has been given and what statements have been made affecting him: and then he must be given a fair opportunity to correct or contradict them."”
Lord Dyson significantly recognised a necessary, but limited, qualification to this principle, at [64]:
“… where the whole object of the proceedings is to protect a commercial interest, full disclosure may not be possible if it would render the proceedings futile. This problem occurs in intellectual property proceedings. It is commonplace to deal with the issue of disclosure by establishing "confidentiality rings" of persons who may see certain confidential material which is withheld from one or more of the parties to the litigation at least in its initial stages. Such claims by their very nature raise special problems which require exceptional solutions. I am not aware of a case in which a court has approved a trial of such a case proceeding in circumstances where one party was denied access to evidence which was being relied on at the trial by the other party.”
The reference there to the special problems raised by intellectual property proceedings may, in my view, similarly apply to competition law proceedings where rival commercial interests are involved.
The application of such protection in intellectual property cases has subsequently been considered by judges of the Patents Court, both before and after the Al Rawi case. In Roussel Uclaf v Imperial Chemical Industries plc [1990] RPC 45, Aldous J explained, at 48:
“In patent actions it is not unusual that documents disclosed on discovery include matters which a party considers contain valuable confidential information. The procedure normally adopted is that disclosure is first made in confidence to counsel, solicitors, independent patent agents and independent experts. In many cases this enables the parties to prepare and argue their cases properly. Further, the parties can often agree that all that is necessary to preserve the confidential information is that certain parts of the documents are blanked out.”
However, in that case after disclosure had taken place the plaintiffs sought to have two patent attorneys working in their patent department admitted to the confidentiality ring, which the defendants strongly opposed. Aldous J stated (at 49):
"Each case has to be decided on its own facts and the broad principle must be that the court has the task of deciding how justice can be achieved taking into account the rights and needs of the parties. The object to be achieved is that the applicant should have as full a degree of disclosure as will be consistent with adequate protection of the secret. In so doing, the court will be careful not to expose a party to any unnecessary risk of its trade secrets leaking to or being used by competitors. What is necessary or unnecessary will depend upon the nature of the secret, the position of the parties and the extent of the disclosure ordered. However, it would be exceptional to prevent a party from access to information which would play a substantial part in the case as such would mean that the party would be unable to hear a substantial part of the case, would be unable to understand the reasons for the advice given to him and, in some cases, the reasons for the judgment. Thus what disclosure is necessary entails not only practical matters arising in the conduct of the case but also the general position that a party should know the case he has to meet, should hear matters given in evidence and understand the reasons for the judgment.”
After balancing the various considerations for and against admitting the two attorneys, Aldous J came down in favour of admitting them subject to strengthened undertakings. The appeal against his decision was dismissed by the Court of Appeal.
In IPCom GmbH & Co v HTC Europe Co Ltd [2013] 52 (Pat), Floyd J (as he then was) considered which persons should be entitled to see confidential documents being disclosed in two related proceedings concerning the determination of the appropriate royalty and other FRAND terms of licences under IPCom’s standard essential patent. A previous order of the court had provided for the establishment of a confidentiality club “to include as a minimum: external counsel, experts and one other person elected by each of the parties.” When the court ordered disclosure of some of the licences entered into by the defendants, HTC and Nokia, with third parties, which IPCom argued could serve as comparators, the defendants did not object to these documents being seen by external lawyers and experts but they objected to IPCom’s wish to disclose them to the two individuals who ran IPCom or to a Dr Roman Sedlmaier and Mr Philipp Kahlenberg.
It was not disputed that the licence agreements were regarded as highly confidential and that knowing the terms on which Nokia and HTC were prepared to grant such licences would be of commercial value to IPCom. The judge quoted the above passage from Aldous J’s judgment in Roussel Uclaf and observed, at [21]:
“The court does not normally operate on the basis that a party will wilfully misuse information disclosed to it. But it is recognised that disclosure of information to a party who is or may become involved in collateral commercial activities may place that party in a difficult position where there was a risk of use or disclosure …”
As regards the individuals, Mr Kahlenberg was said to be an external commercial adviser of IPCom who had been involved in developing IPCom’s licensing strategy since 2003. Dr Sedlmaier was an external lawyer working for a German law firm of which the principal was also a member of IPCom, and he had been involved in coordinating IPCom’s legal strategy since 2007. Floyd J stated, at [30]: “It is clear that both individuals are close to IPCom in the sense that they form an important part of IPCom’s licensing team.”
Floyd J referred to the principles set out in Al Rawi and the need to strike a balance. He took account of the fact that although the material was confidential, it was “not at the high end of the scale represented by secret process cases.” The extent to which it was relevant to the issues in the proceedings was at that stage unclear and the litigation was still at an interim stage. But Floyd J significantly added, at [31(iii)]:
“Nevertheless, points of a very broad brush nature have been made in the pleadings about the relevance or lack of it of Nokia's and HTC's licences, a matter on which IPCom have the right to respond.”
Floyd J noted that the case was unusual in the field of patent litigation in that IPCom was a very small company in terms of personnel. He refused to permit inspection of the documents by the two members of IPCom or by Mr Kahlenberg, whom he regarded as being in a very similar position, since there would be real risk that the confidential information would prove of value to them in licensing IPCom’s portfolio of patents. But by contrast he held that inspection by Dr Sedlmaier should be allowed, stating (at [32(v)]):
“Dr Sedlmaier, on the other hand is an external lawyer bound by a professional code of conduct. I accept that he is, as the evidence shows, extremely close to IPCom, and has been involved in commercial negotiations. Whilst that fact is relied on by HTC, Nokia and the interested parties to make a case for his exclusion, it shows also that, if he is included, the prejudice to IPCom is significantly mitigated. To the extent that he is involved in future negotiations he will have to shut out from his mind anything learned from the confidential documents.”
The use of “legal eyes only” [LEO] confidentiality rings was addressed by Henry Carr J in TQ Delta llc v Zyxel Communications UK Ltd [2018] EWHC 1515 (Ch). The judge summarised previous decisions on confidentiality restrictions in the patents field, including in particular the IPCom judgment, and concluded as follows:
“21. In my judgment, the authorities discussed above establish that it is exceptional to limit access to documents in the case to external eyes only, so that no representative from the party which is subject to the restriction can see and understand those documents. An external eyes tier does not require justification for the restriction by reference to individual documents. It enables one party to decide to exclude all representatives of the opposite party from access to any document that it chooses, and places the onus on the party seeking access to apply to court to obtain it. That approach, in my judgment, is wrong in principle.
…
23. However, it is important to emphasise that:
i) parties may choose to agree an external eyes-only tier, as in the Unwired Planet International case [see below];
ii) confidentiality club agreements are often essential in intellectual property cases; which cases require disclosure of confidential information. In such cases, a regime for disclosure which limits access to sensitive documents to specific individuals within one of the parties, in order to protect confidentiality, is now commonplace;
iii) redactions to documents can be made to exclude material which is confidential and irrelevant to the dispute;
iv) external eyes-only access to individual documents of peripheral relevance, whose disclosure would be damaging, may be justified in specific cases; as in the IPCom case;
v) I do not exclude the possibility that in certain exceptional cases, external eyes-only access to specific documents of greater relevance might be justified, at least at an interim stage.
vi) however, in the absence of exceptional circumstances, each party must be able to see and discuss with its lawyers the relevant parts of the key documents in the case.
24. An external eyes-only tier enables a blanket exclusion of access by one of the parties to the relevant parts of key documents. This is incompatible with the right to a fair hearing under Article 6 of European Convention on Human Rights, and with the principles of natural justice. It is incompatible with the obligations of lawyers to their clients. The principles on which solicitors are obliged to act on behalf of clients instructing them require the sharing of all relevant information of which they are aware.”
Although not cited to Henry Carr J in TQ Delta, or indeed in argument before me, it is appropriate to refer also to the judgment of David Richards J (as he then was) in McKillen v Misland (Cyprus) Investments Ltd [2012] EWHC 1158 (Ch), where the judge noted that restricted confidentiality regimes are not uncommon in intellectual property cases prior to actual trial, but after citing from Roussel Uclaf and considering Lord Dyson’s judgment in Al Rawi, he stated, at [50]:
“In the light of the decision and discussion in Al Rawi, it is my view that at common law the court has no jurisdiction to deny a party access to the evidence at trial. But if the jurisdiction does exist, it is in my judgment so exceptional as to be of largely theoretical interest only.”
McKillen concerned an unfair prejudice petition under the Companies Act 2006 and the confidential information at issue was the personal financial information of the petitioner. It was thus far removed from the situation of a dispute between competitors which arises in intellectual property and competition litigation.
I observe that in a number of competition cases, and also in intellectual property cases, some parts of the evidence at the final hearing have remained subject to a confidentiality ring of lawyers and experts. In Lilly Icos Ltd v Pfizer Ltd [2002] EWCA Civ 2, which it appears was not cited in McKillen or TQ Delta, the Court of Appeal (including Aldous LJ) noted without disapproval that the patent trial had been conducted with some of the documents restricted to a confidentiality ring that included “almost entirely professional advisers”, and the Court recognised the particular considerations which apply in patent cases: see at [25(vi)]. More recently, in Unwired Planet International Ltd v Huawei Technologies Co Ltd [2017] EWHC 711 (Pat), which like IPCom concerned FRAND licences, Birss J explained in his judgment at [25]:
“One of the challenges in trying this case was confidentiality. The arguments, evidence and disclosure documents included a large amount of material in which confidentiality was claimed. Some of the claims were from parties or companies who had been parties (Ericsson and Samsung) but some of the confidential material was confidential to third parties such as licensees. The legal representatives of all parties were privy to all the material but some aspects, e.g. material relating to Samsung or Ericsson, was maintained as confidential from Huawei or Unwired Planet staff. Attempting to determine the confidential status of material during the hearing would have been impossible, so the trial was conducted accepting many of the claims to confidence for the time being. While wide claims to confidentiality had been made before trial, they were reduced considerably at and during the hearing. Thus much of the trial took place in public.”
Similarly, in competition cases, I think that restrictions of this kind may sometimes be justified, and when properly confined I do not believe this has caused problems in the lawyers getting instructions from their clients. In the Pay TV (No 1) appeals before the Competition Appeal Tribunal (the “CAT”), concerning the terms of licences to be granted by British Sky Broadcasting Ltd of sports programmes, although the appeals were against a decision of the Office of Communications (“OFCOM”) some of the appellants and interveners were in effect on opposing sides and confidentiality rings restricted to lawyers and experts were ordered; the only issue which arose concerned whether in-house lawyers could be included: see Virgin Media Inc and ors v OFCOM [2010] CAT 16. Similarly, confidentiality rings comprising only external lawyers and experts were established and applied in the recent CAT appeal in Viasat UK Ltd v OFCOM [2018] CAT 18,where the appellant challenged the decision of OFCOM to authorise Immarsat to provide a system for internet or broadband interconnections to aircraft. The confidentiality regime prevented Viasat and Immarsat, respectively, from seeing each other’s confidential information. I note that in Hollander, Documentary Evidence (13th edn, 2018), the learned editor observes at para 10-10, that the position could hardly be otherwise where the parties are business competitors.
Generally, the parties concur in these arrangements, although they still require the approval of the court, having regard to the principle of open justice. Since the hearing of Google’s pending application is not the trial and Foundem’s application concerns the admission to the ring of an outside expert not individuals from the parties, it is unnecessary for present purposes to decide whether the observations of David Richards J require some qualification. In my view, the important points to emerge from the authorities are that: (i) such arrangements are exceptional; (ii) they must be limited to the narrowest extent possible; and (iii) they require careful scrutiny by the court to ensure that there is no resulting unfairness. Any dispute over admission of an individual to the ring must be determined on the particular circumstances of the case.
The present case
In the present case, although there was some argument between the parties about the previous designation as LEO of some documents which were subsequently redesignated (and Foundem maintained that there continues to be over-designation of documents as confidential, LEO or RLEO), Foundem’s present application was not advanced on the basis that any of the limited number of documents which Google now designates as LEO or RLEO are not highly confidential (although in his letter Mr Klöckner surmises that some of the information they contain may be widely known in the search industry). Google has indeed referred to its algorithms as its “crown jewels”. Equally, the assertion by Ms Ford that these documents are of a technical nature, such that Foundem’s external lawyers cannot properly understand them without expert assistance, was not challenged. That expertise could be provided by Mr and Mrs Raff, but as the directing minds of Foundem they are outside the LEO and RLEO rings and Foundem does not now seek to have them included.
Mr Klöckner has explained in his letter of 11 November 2019 the reasons for wishing to see the restricted documents as follows:
“Google’s LEO and RLEO redactions make it impossible for me to understand the meaning and implications of documents which, Google says, are relevant to the issues in dispute in the Proceedings (which includes documents in addition to those expressly relied upon by Google in its Application). In particular, these redactions mean that I am unable to verify various propositions contained in Mr Kwok’s first witness statement and, if appropriate, to challenge his opinions on the application of Google’s algorithms to Foundem’s website. For example, on the basis of the confidential redacted versions of the LEO and RLEO documents which I have reviewed, I am unable to assess:
A) whether Algorithm A and Panda targeted thin affiliate sites/search in search sites only, or whether they targeted competing vertical search sites;
B) if the signals assigned to these algorithms are suitable to identify thin affiliate/search in search sites, while minimizing false positives or damage to competing sites;
C) if similar or the same rules could be applied to Google’s products;
D) whether the specific set of signals in fact correlate with good user experience or discriminate against low quality results;
E) whether a reasonable assessment of the signals and algorithm(s) would or could have given any indication of any potential anti-competitive consequences of their deployment;
F) whether these signals or guidelines have been publicly available to the industry/community;
G) how transparent Google was about applicable guidelines, potential violations and characteristics of websites it deemed “low quality”;
H) whether Foundem or other parties had a chance to know “what hit them” and whether they could have taken any countermeasures;
I) whether it would be fair to assume that Foundem was aware of any violation or circumvention of guidelines; and
J) whether other search engines were generally using a similar set of signals to ensure search quality.”
In reaching my decision, I have particular regard to the following matters:
Despite the lapse of time since this action was started, because of the protracted EU proceedings it is still at a relatively early stage. However, although Google’s pending application may technically be an interim application, as it seeks to strike out or obtain summary judgment on Foundem’s stand-alone case, it has the potential finally to dispose of a significant part of these proceedings.
The RLEO material, in particular, remain highly confidential and Google is legitimately concerned to prevent it becoming known to Foundem or used by
third parties for the purpose of SEO. Despite the lapse of time, it can fairly be regarded as of greater sensitivity than the licence agreements considered in the IPCom case.
Google’s application is not arguing that there is no cause of action or no arguable case on the basis of the pleadings alone. Nor is Google arguing simply that Foundem’s case is bound to fail because it made the same complaints to the Commission which then did not find that those matters constituted an abuse but based the Decision on narrower grounds. That would be an issue of law on which no evidence is required. Google has chosen to put in and rely on significant evidence, including confidential evidence from what are in effect two technical witnesses, exhibiting documents containing technical details, in support of its application.
In order properly to understand that evidence, Foundem’s lawyers need expert assistance. To adopt what was said in Al Rawi, without such assistance Foundem’s lawyers will not be in a position meaningfully to make submissions on that evidence or contest what is said about the RLEO documents. Moreover, having read Mr Kwok’s evidence, it is clear to me that his statement is not limited to explaining what was provided to the Commission but seeks to explain the operation and aims of Google’s ranking algorithms, and how they have been applied to shopping comparison sites generally and Foundem in particular. Mr Pohl’s witness statement performs a similar function more specifically as regards Google’s AdWords program, which is the subject of claims (e) and (f) as summarised at para 5 above.
Foundem is similarly entitled, for the purpose of resisting Google’s application, to consider the other evidence disclosed by Google to ascertain if that material may put the arguments being advanced by Google about the operation of its algorithms as regards comparison websites in a different light: this therefore covers the other LEO documents. While Mr Klöckner may not need to see those documents for all the purposes he describes, I accept that he reasonably needs to see them for many of those purposes.
Although the Commission may not have needed outside expert assistance to reach a view on Foundem’s complaint, I do not know what internal technical resources and expertise the Commission has available and it manifestly cannot be compared to a firm of English solicitors and counsel. Nor is it clear how far the Commission investigated those aspects of Foundem’s complaints or whether it chose rather to concentrate its attention on the favourable treatment given by Google to its own shopping comparison website.
As regards the identity of Foundem’s expert, Ms Hannah of Foundem’s solicitors explains that it took some time to identify and engage an industry expert with relevant and appropriate expertise in the field of online marketing and search optimisation. She states:
“5.3 …, due to the dependency of many experts in the field of online marketing and search engine optimisation on maintaining a positive, working relationship with Google, all of the other potential expert candidates that we approached stated that they were unwilling to be instructed as experts by a litigant that is pursuing a claim adverse to Google. That is so even for academics.
5.4 Mr Klöckner is the only candidate that we have been able to identify with the requisite qualifications who is willing to be instructed as an expert in legal proceedings against Google….”
I do not find that evidence surprising. Although Google’s solicitors have suggested in correspondence two other individuals who they suggest would be suitable experts – and to whom they would not object – it is fundamental that a party in litigation is not bound to appoint as its own expert someone proposed by the other side.
As regards Mr Klöckner specifically, I think it is regrettable that the evidence for Google sometimes seeks to describe him as a “trade witness”. He is being instructed as an expert and, from what I have seen, he seems to have expertise in what, as appears is common ground, is a very specialised and technical field. He is wholly independent of Foundem. As noted above, Google expressly does not suggest that he would not seek in good faith to abide by the undertaking he would be required to give to this Court. I of course recognise that once information has been acquired, it is inevitably difficult not to use it in another context, but I think his position in that regard is no different from that of Dr Sedlmaier in the IPCom case. Mr Klöckner will have to consider what he conscientiously can or cannot do on other SEO matters having regard to what he may learn from these documents. Ultimately, I consider that the interests of the administration of justice and Foundem’s right to a fair hearing of this case must prevail.
Accordingly, on the basis of Google’s pending application, I consider that Mr Klöckner should be entitled to inspect the LEO and RLEO documents, against an appropriate confidentiality undertaking to the Court. Google did not seek to make any representations on the terms of the undertakings set out in the schedules to the Confidentiality Orders of 26 September 2013 and 13 June 2019. But as Mr Klöckner is, I believe, resident in Germany, I consider that he should in addition undertake that he consents to the jurisdiction of this Court for the purpose of enforcement of his undertakings. I am not thereby suggesting that any enforcement will be necessary, but I think that Google is entitled to this measure of additional protection.
However, there is an alternative course which I consider is available to Google. The evidence of Ms Lawrance does not seem to rely on Google’s LEO or RLEO documents, save only for the final sentence of para 71(a) and cross-references at the end of para 72 and (in general terms) in para 74. I think that sentence could be removed, and the cross-references are unnecessary. Subject to those small alterations, continuing reliance on Ms Lawrance’s evidence alone does not, in my judgment, mean that the desire of Foundem for Mr Klöckner to see the restricted documents should outweigh Google’s concern to protect its business secrets.
The evidence of Mr Kwok and, to a lesser extent, Mr Pohl is more problematic. In the draft of this judgment sent to the parties I had indicated that to avoid an order that Mr Klöckner is entitled to see the LEO/RLEO documents, Google should renounce reliance on their evidence for the purpose of its pending application. In response to that draft, Google’s solicitors submitted that this stricture should apply only to those
parts of their evidence which rely on LEO/RLEO documents, and that Google is particularly concerned to adduce Mr Kwok’s comments on Foundem’s particulars giving evidence of its distinct allegation at para 68A of the RRAPOC, which concerns claim (c) as summarised at para 5 above. Foundem’s solicitors responded email, resisting this approach and urging that the statements of Mr Kwok and Mr Pohl should be withdrawn altogether.
I have therefore re-read the evidence of Mr Kwok and Mr Pohl in the light of the comments from both sides. The difficulty arises because of the point made at para
45.e) above, which is in effect also made in the email from Foundem’s solicitors. As regards Mr Kwok, it seems to me that sections A, D and E of his witness statement, subject to omission of any references they may contain to LEO/RLEO documents, should create no unfairness to Foundem and its expert, who is now in the confidentiality ring, and that on balance they do not justify disclosure to him of the more restricted documents. But, in my judgment, Google should not be able to rely on sections B-C of his statement without disclosure of all the LEO/RLEO documents to Mr Klöckner. As regards Mr Pohl, part of the difficulty is that two of the exhibits were re-designated by Google from LEO to confidential so late in the day that Mr Klöckner did not have the opportunity to consider them before writing about his concerns. However, I note that Mr Klöckner’s letter, from which I have quoted above, principally expresses concern about Mr Kwok’s evidence. Accordingly, Google needs to withdraw only those parts of Mr Pohl’s evidence that rely on LEO/RLEO documents and not the rest of his statement, in order to rely on that evidence in support of its pending application. In reaching this conclusion, I bear in mind the high hurdle which Google has in any event to surmount in seeking to strike out or obtain summary judgment at this stage.
Therefore, I think it is appropriate to give Google a reasonable time to consider what course it wishes to take. If it renounces reliance for its application on its evidence in accordance with paras 50-52 above, then I will not now order that Mr Klöckner be admitted to the LEO and RLEO rings. I should make clear that this means that Google simply abandons reliance on those parts of the evidence, not that it furnishes recast or alternative evidence. But if Google maintains its present course, then for the reasons I have explained I will grant the application that Mr Klöckner be admitted to those two rings until further order. Further, Google is not permitted to serve recast or alternative evidence without the permission of the Court; and if it should apply to do so, Foundem then has permission to renew its application on the basis of any such new evidence served by Google.
Contrary to the submissions of Ms Ford, I do not think it is appropriate at this stage to address the position for the rest of the action and as regards the more extensive disclosure that Google may have to provide in the future. For the reasons explained above, trial is still a long way off and disclosure is far from complete. The shape of these proceedings may change significantly. If Google now abandons reliance on the technical evidence relating to LEO/RLEO documents for its pending application and that application should succeed, then the stand-alone claims drop out of the action and it seems likely that many of those documents will no longer be relevant. Further, the outcome of Google’s appeal in the EU Courts may have a substantial effect on the future of the action. If Google’s strike out application fails, then the Court will then
have an opportunity to address the future conduct of the proceedings and the need to protect confidentiality going forward.
I will consider representations from the parties as to what is a reasonable time to allow Google pursuant to para 533 above.
Postscript
Ms Ford made strong complaint about what she asserted were excessive and unreasonable claims of confidentiality made by Google, through its solicitors, which were then progressively reduced in response to requests and protests by Foundem. Mr Turner vigorously rejected any suggestion that Google or its advisors had behaved improperly and pointed out that Foundem, for its part, had made extensive designations of confidentiality in its own disclosure.
It is neither necessary nor appropriate for me to comment in this judgment on the particular conduct of the parties in this case and, quite rightly, I was not taken through all the correspondence in which these matters were contested. However, I find that there is an increasing tendency for excessive confidentiality claims to be asserted over documents and information in competition law proceedings, only for those claims to be curtailed or renounced in response to protests from the other side or intervention by the court. It is my understanding that the same is the case in intellectual property proceedings. This is wasteful of time and costs, and it is not the way modern litigation should be conducted.
There are of course legitimate trade secrets and currently confidential information that merit protection. But the parties and their advisors should appreciate that redactions from documents on confidentiality grounds prior to inspection and any restriction on inspection to a confidentiality ring are exceptions to the normal regime for disclosure and inspection of relevant documents. In CMCS Common Market Commercial Services AVV v Taylor [2011] EWHC 324 (Ch), Briggs J (as he then was) stated, at [40]:
“In my judgment there is no difference in principle between the ambit of the solicitor's duty, on the one hand, in the conduct and supervision of disclosure and, on the other hand, in the conduct and supervision of any redaction of disclosable documents before they are offered for inspection. Listing documents for the purposes of disclosure and making them available for inspection are both parts of the process more generally called disclosure, and the court is heavily reliant upon the solicitor's duty to carry out or at least personally to supervise both tasks ….”
The decision as to whether confidentiality should be claimed for a document ultimately rests with the client, subject of course to the potential for determination by the court. But just as solicitors will not unquestioningly accept their client’s view as to which documents are relevant for disclosure, I consider that they should not necessarily be satisfied by their client’s view that open inspection of a document should be restricted on confidentiality grounds. Solicitors should advise their client as to the proper limits of confidentiality, given the protection for all disclosed documents
under CPR rule 31.22, and the guidance as to the likely extent of justifiable confidentiality given by the EU Courts: e.g. see the judgment of the Grand Chamber of the Court of Justice in Case C-162/15P Evonik Degussa v Commission, EU:C:2017:205, at paras 64-66 (rebuttable presumption that documents at least five years old have lost their secret or confidential nature). If solicitors have reasonable grounds for supposing that their client has made excessive confidentiality claims, they should investigate the matter carefully and discuss it with their client. The obligations of solicitors in that regard are well summarised in Matthews and Malek, Disclosure (5th edn, 2017), chapter 18.