The CJEU decision

Sky applied for large numbers of trade marks for a very broad range of products, including products far removed from its core business, such as whips, furs or fire extinguishers. This practice made Arnold J (as he then was) conclude that "the reason for including such goods and services was that Sky had a strategy of seeking very broad protection of the Trade Marks regardless of whether it was commercially justified" [par. 250 of the referring decision], and he wanted to know the limits set to such strategies by EU law.

The CJEU called to mind its recent decision in C-104/18 Koton and reiterated that the concept of bad faith "presupposes the presence of a dishonest state of mind". In Koton, the CJEU held that a finding of bad faith is not limited to situations where the applicant had knowledge of use of a similar or identical trade mark by a third party [par. 55]. Rather, bad faith must be established in light of "all the relevant factual circumstances as they appeared at the time that the application was filed" [par. 59], including the "commercial logic" of the registration [par. 62].

Building on this in Sky, the CJEU held that a trade mark application "without any intention to use it in relation to the goods and services covered by that registration may constitute bad faith, where there is no rationale for the application for registration" in light of the aims of trade mark law [par. 77]. However, the mere absence of an intention to use is not enough. Rather, to establish that an unused trade mark was registered in bad faith, it must be shown that there existed (a) a dishonest intention of undermining the interests of third parties, or (b) an intention to obtain an exclusive rights for purposes falling outside the functions of the trade mark [par 75].

Where such a finding is made, bad faith will arise only for the goods concerned, so that a partial finding of bad faith is possible [see also Koton at par. 61].

Lastly, the CJEU held that Section 32(3) of the UK Trade Marks Act 1994 is, in principle, compatible with EU trade mark law. This provision requires applicants to state that the trade mark is being used, or that the applicant has a bona fide intention to use the trade mark. The CJEU held that if this is a "mere procedural requirement relating to the registration" of a mark, it is acceptable [par. 86].

But what can be the effect of this statement? It seems that, under English law, a finding of bad faith is likely where the statement turns out to have been falsely made [see the discussion of the case law in the referring decision, pars. 209-233]. But the CJEU seems to reject this possibility: Member States are precluded from introducing grounds of refusal or invalidity not mentioned in the EU Trade Mark Directive or Regulation. Thus, while the infringement of an obligation to make the use statement "may constitute evidence for the purposes of establishing bad faith", the infringement in and of itself cannot "constitute a ground for invalidity of the trade mark concerned" [par. 86].

Lucrezia has no intention to use her study materials according to their function

Bad faith as a safety valve for the trade mark system…

Insofar as it concerns the issue of bad faith, the CJEU's decision in Sky came as no surprise to this Kat. Since Koton, we have known that bad faith must be given a broad application, in the sense that it can potentially cover any trade mark registration made with a "dishonest state of mind". The decisions in Koton and Sky pave the way for increased use of bad faith arguments in case of abuse of the trade mark system, regardless of the form it takes.Thus, in my post on the opinion of Advocate General (AG) Tanchev, I suggested that his approach to bad faith was sensitive to the issue of deliberately abusive filings. The so-called "Gleissner Files" offer a notorious example: they describe in detail the various practices of a Mr. Michael Gleissner, which include filing hundreds of trade mark applications covering popular names and terms as part of a massive blocking strategy. This is apparently a viable business proposition, probably because, whilst in theory there are no limits to signs that can constitute trade marks, in practice there is a finite supply of signs that are in actual fact suitable and attractive for use as a trade mark [an argument developed in more detail in a recent study by Professors Beebe and Fromer].

Courts have been able to adequately deal with these kinds of abusive practices by weaponizing the bad faith requirement. For instance, in T-82/14 Copernicus, the General Court (GC) confirmed that utilizing a "priority trap" – whereby trade marks are filed, withdrawn prior to having to pay fees, and then re-filed until a later applicant registers a similar sign – constituted bad faith [par. 148]. Likewise, the United Kingdom Intellectual Property Office (UKIPO) held repeatedly that comparable strategies could constitute bad faith [e.g. here, here and here].

The requirements for bad faith formulated by the CJEU in Sky fit these types of situations like a glove: indeed, "obtaining an exclusive right for purposes other than those falling within the functions of a trade mark" concisely but aptly describes the business models of the likes of Mr. Gleissner. It is hoped that European courts will now make short shrift of such practices.

… but how far does it go?

Sky, however, was about trade mark registrations by a company that holds a widely reputed trade mark – Sky – and holds a strong position on the market in respect for a number of services, mostly related to television broadcasting.

The real question is what kinds of limitations these kinds of parties face in registering their trade mark also for other goods, no doubt with a view of keeping the trade mark register clear of other Sky-registrations and to thus help protect the distinctiveness of its core mark. This is not an uncommon strategy, though Sky perhaps went a bit overboard: the referring decision noted that two of the registrations invoked by Sky were over 8 000 words long.

In my view, it will be much more difficult to prove bad faith for these kinds of parties. Perhaps Sky did not intend to ever use its mark for whips, furs or fire extinguishers. But that is not enough, the CJEU tells us. In addition, the applicant must aim to undermine the interests of third parties or seek to obtain exclusive rights not justified by the functions of the trade mark system.

My feeling is that it will not be very easy to establish this in cases where the trade mark is used and has a strong reputation for a wide range of goods, since it will often be legitimate for these kinds of parties to protect their brand also in ancillary markets. Then again, invoking several 8 000+ word trade mark registrations against a company that offers services quite far removed from Sky's core business won't win much sympathy from most courts, and it certainly did not in the referring court. Those courts certainly can read in the CJEU's decision encouragement to severely trim down such registrations.

Will a statement as required by Section 32(3) UK Trade Marks Act 1994 help? Most likely, no. If a trade mark is not used after five years, it is clear that no intention to use existed or was lived up to. In that case, it is unclear what the added value is of a statement that such an intention existed at the time of registration if that cannot by itself establish bad faith. After the Brexit transition period, UK courts may become effectively free to disregard the decision in Sky so they may attach more severe consequences to falsely stated intentions to use. But in the rest of Europe, this Kat expects comparable restrictions will do little to reign in applicants like Sky.

All in all, Sky does not really seem to empower courts to deal with the important issues the case raised. As noted last week, the CJEU's solution to the lack of clarity issue is not really satisfactory. The bad faith issue is left sufficiently open for courts to potentially act against overbroad trade mark applications like Sky's; but it will not be easy to determine a clear boundary. The good part is that Sky seems to be just the beginning, so that exciting developments on these issues no doubt await us.

Because the decision has much to offer, the analysis will proceed in two parts. This post concerns the issue of trade mark registrations suffering from a lack of clarity. A follow-up post will be dedicated to the issue of bad faith registrations in the absence of an intention to use.

The CJEU decision

The first two questions referred by Arnold J (as he then was) were whether a trade mark can be invalidated if one or more of its specified goods lack sufficient clarity and precision to enable third parties to determine the extent of protection sought. The referring court wondered if specifications like "computer software" might be so general, and potentially cover so many different goods, that they may not be compatible with the trade mark's function as an indication of origin.

The CJEU held that the list of absolute grounds of invalidity provided in Directive 89/104 and Regulation 40/94, superseded by but in this respect identical to Directive 2015/2436 and Regulation 2017/1001 (the EU Trade Mark Regulation), is exhaustive. Since lack of clarity is not in the list, it cannot be considered a ground for invalidity under EU trade mark law [par. 58-60].

The CJEU further held that trade mark registrations allegedly lacking clarity cannot be attacked on the basis of the requirement of graphic representation and compatibility with public policy, since those grounds of invalidity do not apply to the types of cases the referring court had in mind [par. 67]. In this respect, it parted ways with AG Tanchev, who had suggested that trade mark registrations lacking sufficient clarity might be contrary to public policy.

At the end of this part of the decision, however, the CJEU suggested a solution: a trade mark "registered for a range of goods and services designated in a manner which lacks clarity and precision is, in any event, capable of being protected only in respect of the goods and services for which it has been put to genuine use" [par. 70]. Thus, the CJEU seemed to indicate that, after expiry of the five-year period within which the trade mark must be put to genuine use, trade marks lacking clarity can be revoked insofar as they have not been put to use.

On its face, the solution offered by the CJEU to the problem of unclear and potentially overbroad trade mark specifications sounds reasonable. But the way this Kat reads it, it raises quite a number of new issues that will require sorting out in future decisions. Thus, it may fall short as an immediate solution to the problem of unclear trade marks.

Only trade marks subject to use requirement

To start with, the solution only helps for trade marks that have become subject to the use requirement of Article 18 EU Trade Mark Regulation. Thus, for a period of five years following registration of a trade mark for a term like "computer software", the trade mark proprietor shall in any event be entitled to "a monopoly of immense breadth which cannot be justified by any legitimate commercial interest of the proprietor" as the referring court put it [par. 171].

That is particularly worrisome where trade marks are "evergreened", i.e. re-registered after expiry of the five-year term. To be sure, the European Union Intellectual Property Office (EUIPO)'s Board of Appeal has recently ruled that such a re-registration, if "manifestly intended to circumvent the obligation to prove use of earlier marks", can be made in bad faith [R 1849/2017-2, par. 82, covered by The IPKat here]. But this line of case law is still being developed and it is not clear that it will always be possible for defendants in infringement suits to rely on it. 

Advocate-General Tanchev's eagerly awaited opinion in case C-371/18 Sky and others was handed down today [see here; prior coverage on this blog here and here]. As Arnold J remarked in his referring decision [here], the case raises important issues of European trade mark law, a sentiment shared by the A-G in the opening paragraphs of the conclusion. Par. 5 captures it nicely:

This [case] raises the question whether we have arrived at a point in trade mark law that grants the trade mark proprietor a position of absolute monopoly in the face of which one can no longer defend himself in infringement proceedings – in spite of the fact that the mark has not been used, and is not likely to be used, for many of the goods and services in respect of which it was registered.

With an introduction like that, it is no surprise that the A-G finds that trade mark registrations made with no intention to use may constitute bad faith, and that goods or services specifications which are vague and uncertain may be contrary to public policy within the meaning of Article 3(1)(f) Directive 89/104 and Article 7(1)(f) Regulation 40/94.

Those findings, if upheld by the Court of Justice EU, could have a significant impact on trade mark law and practice, so let's take a closer look.
The key issues

Sky holds a large portfolio of SKY trademarks which it actively enforces. These registrations cover a very wide range of goods and services: the referring decision notes, in paragraph 4, that two of the trademarks invoked by Sky have specifications of over 8 000 words long (!). Some of these goods, such as computer software, have a very broad scope: after all, every digital product incorporates some kind of software so potential applications of computer software are virtually unlimited.

What is more, many of the goods and services on Sky’s long specifications, such as fire extinguishers, whips and furs, go substantially beyond Sky’s core business, and Sky has no apparent intention of ever using the SKY trademarks for these goods and services.

This gave rise to the two main issues in this case:

1. Are registrations for terms with a potentially very broad scope, such as ‘computer software’, invalid for a lack of sufficient clarity or precision?; and
2. Can it constitute bad faith to apply to register a trade mark without any intention to use it in relation to the specified goods and services?
   
   [Three other questions were also referred, see here]. 

The A-G's opinion

Lack of clarity and precision

As for precision and clarity of trade mark specifications, the A-G concludes that a lack thereof is not, in and of itself, a ground for invalidity. According to the A-G, the grounds of invalidity are listed exhaustively in the Trade Mark Directives and Regulations, and the ground that some or all of the terms in the specification lack sufficient clarity and precision is not among them (paragraph 38-39).

Nor is the CJEU's decision in IP Translator [C-307/10, here] an independent ground for invalidation for lack of clarity or precision. In this respect, the A-G finds that this decision "simply cannot be interpreted as meaning that it introduces a new ground for invalidity" (par. 57).

However, the A-G finds that a lack of clarity and precision may bring a registration within the scope of the public policy exception ("sub (f)" in the various trade mark legislative texts). The A-G concludes that registration of overbroad terms, such as 'computer software', is "unjustified and contrary to the public interest because it confers on the proprietor a monopoly of immense breadth which cannot be justified by any legitimate commercial interest of the proprietor" (par. 79).

To come to this conclusion, the A-G finds that the IP Translator decision allows the inference that unclear and imprecise registrations infringe public order (par. 61); that there are strong policy reasons for refusing unclear and unprecise registrations (pars. 62, 72); and that Sky had no reasonable commercial rationale for seeking such broad registrations, because it did not intend to use the marks for all goods potentially falling within their scope (pars. 66, 68, 71, 73).

Although, in par. 82, the A-G asks the question what the relevant criteria should be for determining sufficient clarity and precision, he does not really develop the point. Rather, he refers to the General Court's case law on partial revocation for non-use under which broad registrations are divided into subcategories and only upheld for subcategories for which use can be shown [e.g. T-126/03, here].

The A-G concludes that this is a sensible approach also in cases where there is no intention to use the mark across the whole breadth of the registration (par. 86). Subcategories for which no such intention exists are liable for revocation on grounds of bad faith, following the A-G's reasoning in the rest of the opinion.

Bad faith in case of no intention to use
As regards the second key issue in the case, the A-G found that in case of a "deliberate intention to acquire rights [the applicant has] no intention to use, potentially, inter alia, to prevent third parties from using the registered mark", then there is bad faith (par. 94).

The A-G notes that a system such as the European system, whereby it is not necessary to demonstrate an intention to use at the time of applying for a mark [contrary to e.g. US trade mark law], has certain advantages: it allows for protection at an earlier stage and makes the system faster, simpler and cheaper. But it also allows the creation of unjustified barriers to use of the sign by others (par. 95).

Thus, it is "relevant" to enquire into the applicant's commercial logic for filing their registration (par. 103). This also follows from the CJEU's decision in KOTON [C104/18 P, here], which the A-G discusses in some detail (pars. 97-101). The A-G also discusses a sample of cases where the filing logic was contrary to the purpose of the trade mark system, for instance in cases of dishonesty, cybersquatting or engaging in a blocking strategy (pars. 107-108, 111-112). There is a growing body of case-law in which the General Court and national authorities have condemned such practices and revoked such registrations as having been made in bad faith.

It seems that the A-G wants to add registrations made without any intention to use to this list, as becomes apparent e.g. from par. 114:

I consider that if the applicant does not intend to use the mark, it is immaterial whether the applicant intends to prevent a specific third party from using the mark, or all third parties. In those circumstances, the applicant is improperly seeking a monopoly to exclude potential competitors from using a sign which he has no intention of using. That amounts to an abuse of the trade mark system.

Comment

The A-G's reasoning shows concern about potential abuses of the trade mark system and rightly so. In light of this, his findings as regards the bad faith issue are not surprising—and they reflect national practice that developed in response to egregious practices, which in some cases have been carried out on a very large scale [e.g. the case of mr. Gleissner, see here].

It is, however, not entirely clear whether for the A-G, the mere absence of an intention to use is enough to conclude bad faith. This seems to follow from some paragraphs (e.g. 84, 110, 114) whereas others suggest some additional requirement, such as a deliberate intention to obstruct third parties (e.g. 94).

In my view, a finding of bad faith on the mere ground that there was no intention to use would contravene a fundamental aspect of European trade mark law according to which an intention to use is not, in fact, necessary for registration of a mark. In real life situations, it can be difficult for companies to prove intention to use and exposing those companies to bad faith claims would be unjustified. It seems that something more is needed, for instance abusive filing patterns such as those at issue in the GC's decision in LUCEO [T-82/14, here], which the A-G also cites.

As for clarity and lack of sufficiency, I agree that there are strong policy reasons to oppose vague and indeterminate registrations. But I have doubts that public policy is the appropriate ground to oppose them. This ground is not intended to oppose trade mark registrations that somehow sit awkwardly with the purpose of the trade mark system and if it were so expanded, many more registrations would potentially come within its ambit, creating great uncertainty.

Instead, the better solution might be just what the A-G proposes in par. 86, whereby such vague registrations may be revoked for subcategories where bad faith exists.