Montres Breguet SA & Ors v Samsung Electronics Co Ltd & Anor

The Claimants are all members of the Swatch group of watch makers. Like the judge below, I shall refer to the Claimants collectively as “Swatch”. Each has a distinct brand and associated trade marks. They increase in price from Swatch in the “basic range”, Tissot, Mido and Hamilton in the “middle range”, Longines in the “high range”, Omega in the “luxury range” to Breguet, Blancpain, Glashütte Original and Jaquet Droz in the most exclusive “prestige range”.

It is common ground that the conventional location of a trade mark on a watch, whether in word or figurative form or both, is generally the centre of the upper portion of the dial. The judge used the term “Dial Branding” to refer to the appearance of the trade mark or sign on the watch face, and I will adopt that expression.

The First Defendant (“Samsung”) is the parent company of the well-known technology group. The claim against the Second Defendant, a United Kingdom subsidiary, was stayed, so Samsung is the only active defendant. Samsung markets a range of smartwatches. Since 2018 its smartwatch models have been called Galaxy.

Swatch claim that between October 2015 and February 2019 23 of their trade marks were infringed by 30 digital watch face applications or “apps” which could be downloaded by users to Samsung smartwatches from the Samsung Galaxy App store (“the SGA store”). These apps had names, and/or when downloaded to a smartwatch displayed signs in the Dial Branding position, which Swatch contend are either identical or similar to one or more of their trade marks. Although the claim only concerns the app names and Dial Branding, many of the apps in issue replicated in meticulous detail the watch faces of genuine watches. An example is show below (genuine Glashütte Original watch on the left, allegedly infringing digital watch face on the right).

Some apps were available for free and others required a modest payment. Samsung has admitted that these apps were downloaded 157,715 times in the European Union as it was then constituted.

Although each of the apps was developed by a third party, Swatch contend that Samsung was intimately involved in, and controlled, the process by which the apps were made available to consumers. Their primary case is that Samsung is liable as a primary tortfeasor. In the alternative, they contend that it is a joint tortfeasor with the app developers. An alternative claim for passing off, covering a further six apps as well, was not pursued at the trial, and claims for infringement of two additional trade marks were also not pursued.

Swatch began monitoring the SGA store in December 2018, when they became aware that potentially infringing apps were available in it. Swatch first made contact with Samsung’s Swiss subsidiary by a letter dated 21 December 2018, which demanded that certain apps, together with any other infringing apps, be taken down, and sought disclosure of the app developers’ details and details of downloads. The letter referred to trade mark infringement, but also made additional allegations, including in respect of copyright in design features. By 9 January 2019 all the apps identified by Swatch had been taken down by Samsung. Swatch commenced these proceedings on 19 February 2019. The claim identified some additional apps not referred to in the correspondence. These were then also promptly taken down by Samsung.

Falk J (as she then was) held for the reasons given in her judgment dated 20 May 2022 [2022] EWHC 1127 (Ch) that most, but not all, of the apps infringed the Swatch trade marks relied on and that Samsung was primarily liable for such infringements. She therefore did not determine Swatch’s alternative claim that Samsung is liable as a joint tortfeasor for infringements committed by the app developers.

The judge granted Samsung permission to appeal on three grounds. First, that she was wrong to hold that the use of the signs complained of was use by Samsung, as opposed to use by the relevant app developer. Secondly, that she was wrong to hold that the use of the signs complained of included use in relation to smartwatches. There is no dispute that there was use, at least by the app developers, in relation to the apps. This ground of appeal goes to the extent of the infringement, if any. Thirdly, that she was wrong to reject Samsung’s defence under regulation 19 of the Electronic Commerce (EC Directive) Regulations 2002, SI 2002/2013 (“the UK e-Commerce Regulations”), which implements Article 14(1) of European Parliament and Council Directive 2000/31/EC of 8 June 2000 (“the e-Commerce Directive”). This defence, if established, is a defence to Swatch’s claims for financial remedies in respect of the alleged infringements. It does not give Samsung a defence to Swatch’s claim for an injunction to restrain future infringements of the trade marks. The judge granted an injunction, and there is no appeal against that decision if the first ground fails.

The judge dealt in detail with each of the trade marks, the apps and the allegations of infringement in three Appendices to her judgment. It is not necessary to condescend to the same level of detail in this judgment because the grounds of appeal are general ones. Samsung does not pursue a number of other defences advanced at trial.

The judge set out the factual background at [39]-[73]. What follows is an abbreviated version of that account.

A smartwatch is a wearable electronic device with a digital interface that provides operational controls through software apps. Smartwatches tell users the time, but also come with a number of additional features not provided by traditional watches, such as the ability to make calls, listen to music, read messages or emails, make payments and track health and activity levels. Thus Samsung smartwatches function more like a wearable computer or mobile phone than like a traditional watch.

The face of a smartwatch consists of a digital display which, unless an “always on” feature is enabled, is switched on when prompted, whether by motion or tapping the screen. When switched on, the default screen is a watch face which tells the time, although features such as incoming calls may be displayed instead. If an “always on” feature is available and enabled, a watch face will be displayed continually, although to preserve battery life and avoid screen damage the version of the watch face shown will generally be a simpler version as compared to the one displayed while the smartwatch is in active use, using around 15% of the number of pixels. If that feature is not enabled the watch face will change to a blank screen after a set number of seconds, the default being 15 seconds.

Samsung smartwatches allow significant flexibility for the consumer to configure the software on their smartwatch as they choose, including by accessing the SGA store to download apps.

During the relevant period a Samsung smartwatch did not connect directly to the internet. Instead, it relied on a wireless connection to a smartphone with which it was linked (not necessarily a Samsung model of a smartphone). In order to create the connection the end user would have to download the Samsung Wear app to the phone.

Since 2015 Samsung smartwatches have been round in shape, have had a bezel and have been fastened to the user’s wrist by a strap. In each of these respects, they are like conventional watches.

Samsung’s smartwatches are, and were at relevant times, specifically marketed by it for their watch-like qualities, including claims, for example, that they are “truly watch-like” and have a “rich watch face design”.

Samsung’s business model involves producing and marketing the smartwatches pre-installed with a comparatively small number of apps designed by Samsung so that the hardware functions. This will include around 20 watch faces that the consumer can choose from. Samsung also offers some further apps that it has developed, including watch face apps, in the SGA store. A substantial majority of apps available in the SGA store are, however, developed and provided by third-party app developers.

Samsung’s evidence was that, based on the available figures, only 26% of UK Samsung smartwatch users had made active use of the SGA store, with those who did downloading an average of 4.6 watch face apps each. About 15% of watch face apps are made available for free, frequently becoming paid-for apps if they gain popularity.

Samsung smartwatches are generally in use for a relatively short period. The average replacement cycle for a Samsung smartwatch in the UK is about 27 months.

The SGA store was established, and is managed, by Samsung specifically for Samsung devices. It is an online platform through which both Samsung and third-party app developers can make apps available to consumers who have Samsung products, including but not limited to Samsung smartwatches.

The SGA store comprises both “back-end” and “front-end” servers. The back-end comprises a developer portal where app developers can upload apps and an administration site where Samsung reviews apps before they are made available. The front-end is the part that is available to consumers.

Also available to developers via the back-end portal is the Galaxy Watch Studio tool. That tool provides user-friendly instructions on how to create Samsung smartwatch watch face apps that work on Samsung’s Linux-based Tizen operating system, for uploading to the SGA store.

Once an app has been designed and uploaded to the SGA store, it is automatically submitted to Samsung’s app review process, which must be passed before it is made available to consumers as a “for sale” version on the front-end server. The review process is discussed further below.

In order to access the SGA store, consumers must set up an SGA store account with Samsung, which involves providing a name, email address, country of residence and phone number (and optionally payment details, although those can be provided when a purchase is made). Watch face apps available in the SGA store can only be viewed or downloaded by consumers with a Samsung smartwatch, which must also be connected to a smartphone (as discussed above). Once connected, watch face apps may be viewed and downloaded to the smartwatch by using either the phone or the smartwatch. The process for accessing the SGA store and obtaining apps is discussed further below.

Although the SGA store has a search function, at least during the relevant period it was restricted to the app name and features such as styles and colours. Moreover, search results were not necessarily comprehensive because unpopular apps might be excluded.

During the period in issue, any third party who wished to upload an app to the SGA store was first required to obtain an account by registering with Samsung’s SGA Seller Portal by providing their name, email address, country of residence and bank account, and agreeing to Samsung’s terms and conditions. The terms and conditions included, among other things:

These terms and conditions were applicable in respect of all apps uploaded to the SGA store, including but not limited to watch face apps.

Samsung has organised developer conferences, which are available online. Their aim includes the provision of guidance about developing apps for the SGA store, with a view to ensuring that a good range of high quality apps is available.

Each third-party app uploaded to the SGA store is, and was at relevant times, subject to both a technical and a content review prior to being made available to consumers. The technical review is limited to checks for viruses and to ensure the app’s functionality and compatibility with Samsung’s operating system. The content review had historically been conducted by a team based in Poland, but in 2016 a new team was set up in the Samsung Vietnam Mobile R&D Centre (“the SVMC”) in Hanoi managed by Loi Le. Since then that team has carried out all content review globally (covering apps for smartphones and tablets as well as watch apps), other than for apps released to the Chinese market.

The team comprises around 14 individuals, all of whom are Vietnamese and were generally recruited as new graduates. Most of the team are software engineers by background. None of them have international or European market experience, and while all have some level of English they are not fluent English speakers. They would have had little knowledge of luxury brands when they were recruited. Mr Le’s own knowledge of famous watch brands at the time appears to have been limited to what he had picked up from sports-related advertising.

Team members are assigned a number of apps each day and are expected to complete their review within three days. They are instructed to spend about 5-10 minutes reviewing each app against Samsung’s internal content review guide (“the CRG”, discussed below), using their own knowledge and judgment and referring to the internet. From January 2018 to March 2019, each member of the team reviewed 85 apps per day on average, working a 44 hour week. That works out at a little over six minutes per app.

The team base their review on metadata, app names, tags, descriptions and screenshots provided by the app developer. They do not ordinarily review the apps as they would appear on a device. There is an ability to escalate borderline cases to be determined by an App Review Team in Korea, but this facility is rarely used.

Once an app has passed the review process it will go live in the SGA store within about two hours, unless the app developer has chosen to distribute it at a later date. The app developer will also determine the price (within permitted parameters) or whether the app is to be available for free.

The content review process has undergone an evolution since the commencement of these proceedings. Prior to February 2019 it was limited to a manual and visual review of app content conducted by Mr Le’s team with reference to the CRG. From February 2019 onwards an additional stage of automated “pre-review” was added to the process in the form of Samsung’s Validation Automation System (“VAS”). VAS has not been used to scan existing apps. It applies only when new apps are reviewed prior to being made available.

VAS is a computerised process that automatically cross checks a proposed app’s meta tags (source code information), title, app description and the developer’s name against a list of keywords, with the aim of identifying content that could be infringing or prohibited. As of February 2022 the database of keywords checked for by VAS comprised approximately 4,000 pre-defined trade marks, brands and phrases. It includes watch brands such as those owned by Swatch. However, VAS only checks for exact matches. Any app with a word or phrase that matches the keywords is flagged for manual review, in addition to the manual content review which continues to apply to all apps.

In addition to VAS, Samsung has also introduced a “Closed-Seller” model for watch face app developers, such that only pre-approved sellers, who have signed a specific pledge not to infringe intellectual property rights, can register to sell apps. It has also introduced an automated visual comparison of watch face app displays against conventional watch faces, although the recognition rate is quite low. Since 2021 app developers have also been required to submit any Dial Branding in the app description, which VAS can then review.

The version of the CRG in evidence is 40 pages in length. It has four sections: (1) performance, which includes functionality and usability; (2) restricted content, which covers inappropriate material such as content that is sexually explicit, violent or drug-related; (3) legal; and (4) culture or market, which covers matters specific to certain countries.

The legal section covers 14 pages, most of which are devoted to intellectual property. There is a detailed section covering the circumstances and context in which references to Samsung’s own trade marks are permitted, with a table of screenshots providing examples as well as relatively detailed guidance. The section devoted to third-party rights holders (referred to as “copyright owners”) contains much less detailed guidance. It provides pictorial examples of a number of well-known sports club logos, cartoon characters, games brands and “copyrighted brands” such as IKEA and McDonald’s, and describes the process for reporting infringing content. No watch brands are included and there is no specific guidance on apps for smartwatches. The substantive guidance amounts to little more than the following:

There was no clear instruction to do an internet search of any kind, at least in respect of what the reviewer did not regard as “well known contents”, and certainly no instruction to check any trade mark database.

Like other online marketplaces, the SGA store operates a “notice and take down” process, whereby anyone can notify Samsung that a particular app infringes intellectual property rights. Samsung will investigate (suspending sales of the app while it does so) and, if appropriate, will remove the app from the store. There was no dispute about the effectiveness of this process at trial.

Access from the smartwatch. Samsung smartwatch owners are first introduced to the SGA store during the set-up phase, with a “welcome” screen. By clicking on the icon the user is taken to a screen entitled “Legal information”, containing terms and conditions which the user must accept in order to proceed, and commencing with the words “All content made available to you by the Samsung Galaxy Store is subject to the terms and conditions set forth below”.

Once accepted, the user is presented with a range of icons on the watch face, one of which is the SGA store. On accessing the store different options are presented, one of which is “Watch Faces”. Within this users can scroll between different watch face apps, or make searches. During this process they will see the name of the app, a rating, a small icon previewing its design and an “install” button allowing for installation directly from that screen. No information about the seller of the app is provided at any stage of the process just described, and Samsung’s own apps and third-party apps are not distinguished. However, by clicking on the app name rather than pressing “install” at this stage a new screen will be shown which additionally includes a previously unseen name below the app name. That name is the name of the app developer, although that fact is not stated anywhere on the screen. Prior to 2019 the first screen did not include the install button, and instead clicking on it led to an installation screen which included the developer name. An example of the first screen from 2018 is shown below.

Access from the smartphone. Users may also search for watch face apps via their smartphones, if connected to a Samsung smartwatch. The phone version of the SGA store app allows users to select from different categories of apps, one of which is dedicated to watches and displays only watch apps. Again, they can scroll or search. In this case the details that will be displayed with the app include the name of the app developer as well as the other details shown on the watch, although again without specifying what the name represents. On the other hand, the facts that (i) more than one app can be previewed at once on a phone, (ii) there is a variety of different names and (iii) on some of them Samsung’s own name might appear instead of a third-party name does more to indicate that the name may relate to the source of the app. Clicking on the app name brings up a further screen which again contains the app developer’s name under the app name. It also separately contains “Seller info”, but that appears not to repeat the app developer’s name and to relate instead to version number and file size.

This claim was commenced, and concerns alleged infringing acts committed, when the UK was a Member State of the EU. Furthermore, most of the trade marks relied upon by Swatch were, at the time of the alleged infringements, EU trade marks. For these reasons it is common ground that the substantive trade mark legislation applicable to most of the acts in question is either Council Regulation 207/2009/EC of 26 February 2009 on the Community trade mark (codified version), 207/2009 as amended with effect from 23 March 2016 by European Parliament and Council Regulation 2015/2424 of 15 December 2015 or European Parliament and Council Regulation 2017/1001/EU of 14 June 2017 on the European Union trade mark (codification). There is no material difference between the respective provisions of the Regulations, and it is convenient to refer solely, as the judge did, to Regulation 2017/1001. Although one of the trade marks which the judge held had been infringed is a UK trade mark, it is common ground that this makes no material difference to the applicable law.

Article 9 of Regulation 2017/1001 provides, so far as relevant:

It is common ground that regulation 19 of the UK e-Commerce Regulations faithfully implements Article 14(1) of the e-Commerce Directive. It is also common ground that, since the e-Commerce Directive includes a number of recitals which assist in the interpretation of Article 14(1) and since Article 14(1) has been interpreted in a number of judgments of the Court of Justice of the European Union, it is more convenient to refer directly to the e-Commerce Directive.

The e-Commerce Directive contains three exemptions from liability (also referred to as “safe harbours”) for intermediary service providers in respect of acts referred to by the captions “mere conduit” (Article 12), “caching” (Article 13) and “hosting” (Article 14). These exemptions are horizontal in effect, that is to say, they are exemptions from any form of liability rather than being solely concerned with infringements of intellectual property rights. We are concerned with the third of these exemptions.

The recitals to the e-Commerce Directive include the following:

Article 14(1) provides:

The exemptions do not prevent national courts or administrative authorities from requiring service providers to terminate or prevent infringements: see Articles 12(2), 13(2) and 14(3). On the other hand, Article 15(1) provides:

These provisions are to be interpreted in accordance with the case law of the Court of Justice of the European Union decided prior to 31 December 2020, which constitutes retained EU law pursuant to the European Union (Withdrawal) Act 2018. Neither side invited this Court to depart from such case law. Furthermore, it is common ground that two decisions of the Grand Chamber of the Court of Justice delivered since 31 December 2020 which refine the Court’s previous case law are, although not binding, of persuasive authority.

It is well established that many questions in trade mark law are to be assessed from the perspective of the average consumer, who is deemed to be reasonably well informed and reasonably circumspect.

The judge held that two classes of consumer were relevant in this case:

Swatch claim that the uses of the signs complained of infringe pursuant to Article 9(2)(a) or (b) or (c) of Regulation 2017/1001. Each of these forms of infringement requires “use” of the sign in question by the person alleged to infringe. It is common ground that each of the signs was used by the respective app developer. Swatch contend the signs were also used by Samsung. Samsung disputes this.

The CJEU has repeatedly held that a party only “uses” a sign for this purpose if it uses that sign “in its own commercial communication”: Joined Cases C-236/08 to C-238/08 Google France SARL v Louis Vuitton Malletier SA [2010] ECR I-2417 (“Google France”) at [56]; Case C-324/09 L’Oréal SA v eBay International AG [2011] ECR I-6011 (“L’Oréal”) at [102]; Case C-379/14 TOP Logistics BV v Bacardi & Company Ltd [EU:C:2015:497] (“TOP Logistics”) at [41]; Case C-567/18 Coty Germany GmbH v Amazon Services Europe SARL [EU:C:2020:267] (“Coty”) at [39]; Joined Cases C-148/21 and C-184/21 Louboutin v Amazon Europe Core Sàrl [EU:C:2022:1016] (“Louboutin”) at [29]. A person may allow its clients to use signs without itself using those signs: Google France at [56]-[57]; L’Oréal at [102]-[104]; Case C-119/10 Frisdranken Industrie Winters BV v Red Bull GmbH [EU:C:2011:837] (“Frisdranken”) at [28]; TOP Logistics at [45]; Coty at [39]; Louboutin at [29]. Merely creating the technical conditions necessary for the use of a sign and being paid for that service is not sufficient to amount to use: Google France at [57]; Frisdranken at [29]; Coty at [43]; Louboutin at [31].

In the recent judgment of the Grand Chamber in Louboutin, which was delivered after the judge’s judgment, the Court noted at [34] that neither L’Oréal nor Coty had addressed the situation where an online sales website incorporated sales offers made by the operator of that site as well as offers by third parties. As the Court explained at [35]-[36], the referring courts had specifically raised the issue of the impact of this, and the significance in that context of the perception of the users of the website in question and other circumstances.

The Court addressed that issue as follows:

The judge set out the law at [84]-[91]. Although she did not have the benefit of Louboutin, neither side criticised her exposition.

The judge set out the parties’ cases on use at [92]-[94]. In this context she recorded Swatch’s reliance upon the decision of John Baldwin QC sitting as a Deputy High Court of Justice in Cosmetic Warriors Ltd v Amazon.co.uk Ltd [2014] EWHC 181 (Ch), [2014] FSR 31, which can now be seen to have anticipated the reasoning of the CJEU in Louboutin.

The judge set out her assessment at [95]-[110]. Her reasoning may be summarised as follows.

Many of the features she had taken into account would not, or would be unlikely to, amount to use if taken in isolation, but Samsung’s conduct must be considered as a whole: [95].

An overarching point was that Samsung marketed its smartwatches as truly watch-like: [96]. Furthermore, Samsung specifically advertised the availability of a wide range of watch face apps in the SGA store with a view to making its products appear more attractive. Samsung encouraged app developers to develop the vast majority of these: [97].

It was in Samsung’s commercial interests to host third-party apps as well as in the interests of app developers: [98]. There was a symbiotic relationship between app developers and Samsung going well beyond the revenue derived from the sale of apps and benefitting both parties: [99]. For this reason Samsung assisted app developers in various ways, such as by means of the Galaxy Watch Studio tool: [100]. The apps in question were designed exclusively for, and operated only on, Samsung smartwatches: [104].

It was relevant that the SGA store was operated by Samsung and dedicated to apps for Samsung products, and that Samsung reviewed all apps for both functionality and content before they were made available in the SGA store. Given these facts and the fact that Samsung had specifically marketed its smartwatches by reference to the availability of watch face apps in its store, apps available in the SGA store would be understood by the average consumer to carry an implicit assurance that Samsung is satisfied with them and ought to provide assistance in the event that they prove problematic: [101].

Watch face apps did not relate to goods or services provided by a third party, they were grouped together in the SGA store with watch face apps designed by Samsung and they provided an optional alternative to the Samsung-designed watch faces preloaded on the watch: [102], [105].

A watch face app could be perceived as a key part of both the cosmetic design and the working mechanism of the smartwatch itself, rather than simply an app that could conveniently be accessed through the watch: [103], [105].

A poor-quality watch face app designed for the Samsung smartwatch and acquired through the SGA store would reflect negatively on Samsung in the mind of at least a significant proportion of average consumers. A clear aim of the content review is to limit the risk of that occurring, and rather to ensure that the apps that make it through the review are good quality apps that will enhance, rather than diminish, customers’ perceptions of Samsung’s products: [106].

Samsung not only orchestrated the provision of the apps to customers, but also would field customer complaints about them, provide a level of customer support and share in any revenue raised: [107].

The fact that, once an app had passed Samsung’s review, the developer could determine when it became available to consumers, and (within limits) the price, was insufficient to alter the conclusion that there was use by Samsung: [108].

Samsung’s role went well beyond that of the can filler in Frisdranken, the warehouse-keeper in Coty, the online marketplace in L’Oréal and the search engine provider in Google France. It was not just providing the necessary technical environment, but had a clear commercial interest in the watch face apps in the SGA store, their availability, presentation and use by consumers: [109].

Thus Samsung had used the signs in its own commercial communications: [110].

The judge’s decision on this issue involved a multi-factorial evaluation. It can therefore only be disturbed on the grounds explained in Re Sprintroom Ltd [2019] EWCA Civ 932, [2019] BCC 1031 at [72]-[78] (McCombe, Leggatt and Rose LJJ).

Samsung contends that the judge erred because she took what Samsung describe as an “overly broad” approach. By this Samsung means that she took into account no less than eight factors which it claims are irrelevant. Samsung also complains that she failed to take into account three factors which are relevant.

I do not propose to go through each of the eight factors which Samsung contends are irrelevant since they cover all of the judge’s reasoning I have summarised above. Their alleged irrelevance is based upon two key propositions. First, Samsung says that the judge failed to appreciate that it had done no more than to create the technical conditions for the use of the signs, allow use of the signs and (in most cases) receive payment. The factors she relied on were inherent in the operation of an app store and had nothing to with the immediate context of the use of the signs. Secondly, Samsung says that the judge was wrong to take into account matters which the average consumer would not be aware of and therefore would not cause the consumer to form a link between the sign and Samsung.

I do not accept either of these propositions. So far as the first is concerned, as the judge’s careful analysis demonstrates, Samsung’s actions went well beyond merely creating the technical conditions for use of the signs, allowing use of the signs and receiving payment. The factors she relied on were not inherent in the operation of an app store and they did bear upon the context in which the signs were used. It is not necessary to confine attention to the “immediate” context of the use, whatever that means: see Specsavers International Healthcare Ltd v Asda Stores Ltd [2012] EWCA Civ 24. [2012] ETMR 17 at [87] (Kitchin LJ, with whom Sir John Thomas PQBD and Black LJ agreed).

As for the second proposition, in reaching her conclusion the judge was not merely entitled, but correct, to take into account matters which, even if they were in themselves unknown to consumers, nevertheless affected consumers’ perception of the signs: see Louboutin at [49] and [53]. It would be different if the judge had relied on matters which could not have affected consumers’ perceptions, but she did not. Not having the benefit of Louboutin, she did not in terms ask herself whether consumers would establish a link between Samsung’s service and the signs in question, but her approach was entirely consistent with that test. In particular, she expressly considered the presentation of the disputed apps in the SGA store at [96]-[97], [101]-[102], [105] and [109], taking into account factors such as the grouping together of the Samsung-developed and the independently-developed apps. Furthermore, she expressly considered the average consumer’s perception of the signs at [101] and [105]-[106] and this is implicit in her reasoning at [96]-[97], [103] and [107].

Turning to Samsung’s argument that the judge failed to take into account relevant factors, the first is that the price of the app would be set by the developer. This was a factor the judge expressly took into account, however, at [108]. The weight to be given to this factor was a matter for her. In any event, I am wholly unpersuaded that the judge was wrong not to give this more weight. There is no reason to think that consumers would be aware of this: all they would be aware of was the fact that apps were available at a variety of prices (including a few free ones). That would not affect consumers’ perception that Samsung was either the origin of, or had at least some responsibility going beyond that of a mere retailer for the quality of, the apps supplied under and by reference to the signs.

The second factor is that Samsung discouraged app developers from infringing the intellectual property rights of others. This is not a factor which the judge explicitly took into account in this context, although she mentioned it elsewhere in her judgment. Again, however, I am not persuaded that the judge was in error. This is not something that consumers would necessarily be aware of, and even if they were it would not affect their perception of the signs.

Thirdly, Samsung contends that the judge failed to consider each of the uses individually, and specifically how and in what context the customer would encounter the signs (the so-called “customer journey”). This criticism is unfair to the judge. She did consider each of the uses individually in her Appendices, she did consider the customer journey (or journeys, since, as she explained, the journey differs depending on whether the customer is selecting the app from their smartwatch or their phone) and she did consider how and in what context the customer would encounter the signs. Samsung argues that a key factor is that, when downloading the app from the phone, the consumer would also see the developer’s name. There are two problems with this argument. First, as the judge found at [73] and [154], even if the name was seen, the consumer would not necessarily know that it was the name of the developer (i.e. indicative of a trade origin other than Samsung) rather than, say, a designer or design house (which might be thought to have been commissioned by Samsung). Secondly, as the judge found at [72] and [154], a significant proportion of consumers would download the app from their smartwatches without seeing the name or appreciating its significance if they did.

In my judgment none of Samsung’s criticisms demonstrate any error on the part of the judge, and she was fully entitled to reach the conclusion she did.

Each of the forms of infringement alleged by Samsung requires use of the sign in question “in relation to” relevant goods or services. There is now no dispute that all the signs complained of were used in relation to the respective apps. Swatch contend that they were also used in relation to smartwatches. The basis for this contention is that, once a watch face app has been downloaded and installed on the user’s smartwatch, the sign complained of will appear in the Dial Branding position. Swatch accept that the user will not rely upon the sign as denoting the origin of the smartwatch (which the user will know perfectly well is a Samsung smartwatch), but contend that some third parties (such as friends and acquaintances of the user) will do. Samsung disputes this.

Use of a sign “in relation to” goods or services means use “for the purpose of distinguishing” the goods or services in question, that is to say, as a trade mark as such: see Case C-63/97 Bayerische Motorenwerke AG v Deenik [1999] ECR I-905 at [38], Case C-245/02 Anheuser-Busch Inc v Budĕjovický  Budvar np [2004] I-10989 (“Anheuser-Busch”) at [64] and Case C-17/06 Céline SARL v Céline SA [2007] ECR I-7041 at [20].

Most trade marks are visible, and relied upon by the consumer, at the point of sale. But other trade marks, while visible at the point of sale, are not relied upon by the consumer at that point. An example of this type of trade mark is what is probably the oldest form of trade mark in the world, namely a potter’s mark. A third category of trade mark is not even visible at the point of sale. An old example of this category is the mark traditionally placed upon the cork in a bottle of wine. The fact that a mark is not relied upon, or is invisible, at the point of sale does not mean that it does not function as a trade mark. It still functions as a trade mark because it operates as a badge of origin, and hence quality, after the goods have been sold. It does so not primarily to the purchaser of the goods, who is likely to be aware of their origin, but to third parties who encounter the goods after sale. It is a very old human trait to wish to acquire a product that one has seen worn by a friend or acquaintance or in their home. Furthermore, the goods may be consumed or used by persons other than those who purchased them.

For these reasons, it is well established in both EU and domestic case law that it can be relevant to take the post-sale context into account when considering trade mark issues, including issues as to use: see in particular Case C-206/01 Arsenal Football Club plc v Reed [2002] ECR I-10273 at [57]; Anheuser-Busch at [60]; Case C-361/04 P Ruiz-Picasso v Office for Harmonisation in the Internal Market (Trade Marks and Designs) [2006] ECR I-643 at [48]; Datacard Corp v Eagle Technologies Ltd [2011] EWHC 244 (Pat), [2012] Bus LR 160 at [277]-[289] (Arnold J); Thomas Pink Ltd v Victoria’s Secret UK Ltd [2014] EWHC 2631 (Ch), [2014] ETMR 14 at [142]-[145] (Birss J); and London Taxi Corporation Ltd v Frazer-Nash Research Ltd [2017] EWCA Civ 52 (Ch), [2017] FSR 7 at [34]-[35] (Floyd LJ, with whom Kitchin LJ agreed).

The judge accepted Swatch’s case on this issue for the reasons she gave at [145]:

Samsung does not dispute that, in principle, it is relevant to take the post-sale context into account when considering this issue. Samsung contends that the judge was wrong because there is no realistic likelihood of persons other than the wearer of the smartwatch taking the sign to denote the origin of the smartwatch.

This is essentially an appeal against a finding of fact, but counsel for Samsung did not begin to demonstrate that the judge’s finding was not open to her. He made two points. The first was that, in so far as the judge’s reasoning was based on the watch being in “always on” mode, the simpler version of the watch face using only 15% of the pixels would not display the signs, or at least not legibly. The judge found at [27], however, that a substantial proportion of the apps in dispute did show the brand name in “always on” mode, implicitly legibly. She does not appear to have been asked to make individual findings in relation to each app.

The second point was that, if “always on” mode was not enabled, the sign would only be visible for short periods of time, 15 seconds by default. We were informed by counsel for Swatch, without dissent from counsel for Samsung, that the screen timeout could be set at up to 5 minutes. In my judgment it does not make any difference whether the sign is visible for 15 seconds at a time or 1 minute or 5 minutes. The point of the apps in question, as the judge explained, is to create a realistic simulacrum of a watch face on a smartwatch which is designed by Samsung to be “truly watch-like”. A person who sees a person wearing such a smartwatch may well not know that the watch is made by Samsung. If they are interested to know who made it, the natural place to look is the Dial Branding. The app in question will display a sign identical or similar to one of Swatch’s trade marks. Accordingly, they will be led to believe that this denotes the origin of the watch unless told otherwise. The fact that this may be a relatively infrequent occurrence is neither here nor there.

Accordingly, the judge was not merely entitled to make the finding she did, but also in my view correct.

A subsidiary issue is whether, if the signs were used in relation to smartwatches, that was use by Samsung even if there was use of the signs by Samsung in relation to the apps. The judge addressed this question at [107];

Samsung challenges this conclusion on the ground that the judge was wrong to hold that Samsung affixed the signs to smartwatches. Counsel for Samsung barely pursued this challenge in oral argument. As Swatch point out, and as the judge recognised, the definition of “use” in Article 9(3) of Regulation 2017/1001 is open-ended. It does not matter whether Samsung affixed the signs to smartwatches provided that they used the signs in relation to smartwatches in some way. The judge held at [110] that Samsung used “the signs contained in [the apps]”. In my view she was entitled to reach that conclusion. Imagine that someone is wearing a Samsung smartwatch with a Tissot watch face app. Suppose that a friend says “I like your watch, I see it’s a Tissot” and the wearer replies “No, it’s a Samsung watch with a Tissot face I got from Samsung”. If the wearer thinks that Samsung is the origin of the app, or at least has some responsibility for its quality, they may well convey that impression to their friend.

As explained above, Samsung contends that it has a defence to financial remedies under Article 14(1) of the e-Commerce Directive. Swatch dispute this. There is no dispute that Samsung: (i) provided an “information society service”; (ii) did not have “actual knowledge” of the infringing acts within the first limb of Article 14(1)(a) prior to being notified of them by Swatch; and (iii) acted expeditiously to remove the relevant information upon being notified in accordance with Article 14(1)(b). The judge held that Samsung was “aware of facts or circumstances from which the illegal activity … is apparent” within the second limb of Article 14(1)(a). Samsung challenges this conclusion. Samsung evidently regards this as both its strongest ground of appeal and its most important one, because counsel for Samsung argued it first and spent most time on it both in Samsung’s skeleton argument and in his oral submissions. There is a prior question, however, which is whether Samsung’s acts are within Article 14(1) at all. Swatch contend by a respondent’s notice that they are not. Although strictly speaking it is only necessary to consider a respondent’s notice if the relevant ground of appeal succeeds, logically the issue raised by Swatch comes first. Furthermore, Samsung’s appeal with respect to Article 14(1)(a) raises an important question as to whether the judge placed too great a burden on a party in the position of Samsung to be pro-active in detecting trade mark infringements by its users which should only be addressed if it is necessary to do so.

In Google France the CJEU held as follows:

In L’Oréal the Court of Justice reiterated at [113] what it had said in Google France at [114] and went on:

In Joined Cases C-682/18 and C-683/18 Peterson v Google LLC [EU:C:202:586] Advocate General Saugmandsgaard Øe expressed the opinion at [139] that “while Article 14(1) of Directive 2000/31 is, as a matter of principle, inapplicable where a provider communicates to the public its ‘own’ content, that provision may, on the other hand, be applicable where, as is this case in this instance, the content communicated was provided by the users of its service”.

In its judgment in that case [EU:C:2021:503] the Grand Chamber of the Court of Justice repeated at [105] what it had said in Google France at [112]-[113] and at [106] what it had said in L’Oréal at [113] (and hence what it had said in Google France at [114]). It went on:

Although Peterson is a case about communication to the public of copyright works, and therefore caution must be exercised before transposing the reasoning to the trade mark context, this strongly suggests that, if a platform operator plays a sufficiently active role that it “uses” the signs in question, then it is unlikely to be “merely technical, automatic and passive” with “no knowledge of or control over the content it stores” so as to be within Article 14(1). In other words, in those circumstances the operator will not have merely stored, processed and displayed information provided by a recipient of the service. The only relevant difference between the two tests that either counsel was able to suggest is that “use” must be considered from the perspective of the average consumer of the relevant goods or services, whereas this does not apply to Article 14(1).

The judge’s assessment was as follows:

Nevertheless the judge went on at [222] to say that she preferred to base her conclusion upon Article 14(1)(a).

Swatch contend that it inexorably follows from the judge’s findings on the question of “use” of the signs by Samsung that Samsung’s role was an active one, rather than “merely technical, automatic and passive” with “no knowledge of or control over the content it stores”, and that this is confirmed by her reasoning at [218]-[219].

Samsung advanced two arguments against this contention. The first is that app stores are not excluded from Article 14(1) by (i) the necessary involvement of the operator in determining whether an app could appear in its store or (ii) the operator undertaking a review to check for (and prevent) illegality. This argument is not to the point. Swatch do not suggest that app stores can never benefit from Article 14(1). But equally they are not inevitably covered by it. Article 14(1) is concerned with the acts in issue, not the type of business that has carried out those acts. The question here is whether Samsung’s acts in relation to the particular apps in dispute went beyond the merely technical, automatic and passive. Samsung did not merely decide to make the apps available and check them for illegality.

Samsung’s second argument is that it should not be penalised for undertaking content review in order to try to prevent illegality, rather than simply taking illegal content down when notified of it. On the contrary, it is in the interests of rights holders and in the public interest that it should undertake content review. This is a familiar conundrum with the exemptions in the e-Commerce Directive. The EU legislature has recently attempted to address this in Article 7 of European Parliament and Council Regulation 2022/2065/EU of 19 October 2022 on a Single Market for Digital Services (Digital Services Act), but that legislation does not apply to the UK and would not be applicable to the facts of the present case anyway. As the law stands in this country, the answer to the conundrum is that an intermediary service provider is not obliged to undertake content review and may choose simply to implement notice-and-take down (although this may carry with it an obligation not merely to take down illegal content, but also to keep it down, and the provider may be required to go further still by an injunction, subject to the constraint imposed by Article 15(1)). Many providers wish to undertake content review for their own commercial reasons, however. If they do so, they have to accept the risk that they may not be able to rely upon Article 14(1). In any event, however, in the present case Samsung’s acts went beyond merely making the content available after a content review as I have already discussed.

In my judgment Samsung’s acts of use of the disputed signs were active, and gave it knowledge of and control over that content. They were not merely technical, automatic and passive with no knowledge or control. Thus they were not within Article 14(1).

It is therefore unnecessary to consider whether Samsung’s acts were within the second limb of Article 14(1)(a). It follows that I should not be taken either to endorse or to criticise the judge’s reasoning on that issue. All I will say, in the light of Lewison LJ’s judgment, is that it seems to me to be arguable: (i) that a diligent economic operator must be taken to be aware of facts and circumstances which are well known (which might include the existence of well-known trade marks); and (ii) that awareness of facts and matters indicative of potential illegality requires a diligent economic operator to undertake a basic level of checking (which might include searching the World Intellectual Property Organisation’s Global Brand Database, referred by the judge at [226](d)).

For the reasons given above I would dismiss the appeal.

I agree, subject to my next sentence. On Arnold LJ’s analysis of the appeal, it is not necessary to say anything about the question whether the second limb of Article 14(1)(a) applies. I would therefore prefer to express no view on that question.

I agree with Arnold LJ that this appeal must be dismissed. Although our decision does not depend on it, I wish to add some short observations on article 14.

Arnold LJ has already set out recital (46) of the e-Commerce Directive which bears on article 14. That recital contemplates that the provider of an information society service will act expeditiously “upon obtaining actual knowledge or awareness of illegal activities”. I read that as involving actual knowledge or actual awareness. The defence to liability under article 14 (a) has a condition attached, namely that:

Thus in the case of actual knowledge, it must be actual knowledge of illegal activities. In the case of awareness, it must be actual awareness of facts or circumstances from which the illegal activity is apparent.

Although the phrase does not appear in the legislation itself, the court has introduced the concept of the “diligent economic operator”. In L’Oreal v eBay (to which Arnold LJ has referred) the court addressed this as follows:

The question is whether the provider of the information society service was:

The court concluded at [124]:

This approach, in my judgment, requires the court to assess what facts and circumstances the service provider was actually aware of; and then to ask whether on the basis of those facts and circumstances alone a diligent economic operator should have realised that there was illegality. The illegality must be “specifically established or readily identifiable”: (Joined Cases C-682/18 and C-683/18) Peterson v Google LLC [2021] Bus LR 1196 at [113].

There is, in my view, no warrant for attributing to the service provider awareness of any additional facts or circumstances which would have assisted in deciding whether or not illegal activity was taking place. That, to my mind is consistent with article 15; and with the court’s statement in Peterson (also at [113]) that there is no “general obligation actively to look for facts or circumstances indicating illegal activity”.

In short, the first part of the test is to ask what facts or circumstances the service provider was actually aware of. That is nothing to do with a diligent economic operator. The second part of the test is to ask: on the basis of those facts and circumstances would the illegal activity have been readily identifiable by a diligent economic operator?

In so far as the judge went further, and appeared to prescribe a series of steps that Samsung ought to have taken (but did not) I consider that she was in error.