Emphasising the highly fact-specific and subjective nature of cases under the Trade Marks Act 1995 (Cth) particularly as it pertains to personal names and merchandising activity, Australia’s High Court has split 3:2 in rejecting a challenge to the registered trade mark KATIE PERRY owned by the Australian fashion designer (Designer). The challenge to the registration was made by the famous popstar Katherine Hudson, better known as KATY PERRY (Singer).
By majority, the High Court has restored the outcome at first instance and overturned the decision of the Full Federal Court which would have cancelled the registration. The majority decision is itself split over 3 judgments, from Jagot J, Steward J and Gleeson J, with Gordon ACJ and Beech-Jones J dissenting.
The Justices were unanimous on construction of section 60 of the Act. The dissent from Gordon ACJ and Beech-Jones J under section 60 was premised on a differing application of that law to the facts at hand at the relevant date, as also occurred in respect of section 88(2)(c).
In the result, Jagot J, Steward J and Gleeson J each considered that a consumer encountering KATIE PERRY on clothing as at the priority date of the registration (29 September 2008) or at the date of the challenge to the trade mark registration (20 December 2019) would not be at risk of confusion due to the reputation of the popstar Katy Perry developed in respect of entertainment services and on recording discs used to communicate recorded music. At the later date, the fame of the Singer in fact pointed away from confusion.
Gordon ACJ and Beech-Jones J on the other hand, considered that there was a likelihood of confusion at each date. They agreed with the Full Federal Court that the lack of instances of actual confusion was given undue weight by the primary judge.
All 5 judges agreed that s 88(2)(c) should be assessed by reference to notional use within the specification rather than the fashion designer’s actual use within that broader specification for clothing.
What does the High Court’s decision mean?
This decision underscores the importance of being proactive with your trade mark portfolio, by ensuring all relevant trade marks are registered for goods and services for which there is use or intended use. For other goods and services of strategic relevance from a sword and shield perspective, consideration should be given to filing for defensive trade mark registrations (we have written about them here). Brand owners also need to monitor for concerning third-party filing activity and enforce rights to protect their exclusivity.
There are grounds to oppose third party filings based on reputational rights and bad faith, but the outcome will always be highly fact-specific – is your reputation in relation to your goods or services such that the use of the other trade mark in respect of different goods or services would be likely to deceive or cause confusion? It is not enough to establish that it is common for providers of your goods or services to also provide the different goods or services. As the High Court’s decision illustrates, that will depend on the nature and expected consequences of the specific reputation that has been acquired in Australia in the minds of ordinary Australian people at the relevant date. In this case, the High Court looked to post-priority circumstances, including that KATY PERRY branded clothes included graphics or photographs of the Singer and that there was not ‘a skerrick’ of evidence of actual confusion. This supported the finding of the majority that the ordinary person would expect something more than the use of the trade mark to signify a trade connection.
Even the dissenting judgment emphasised that a reputation does not amount to de facto protection for all goods or services:
‘This conclusion does not mean that all prominent celebrities and entertainers with a reputation that amounts to a mark necessarily acquire some de facto protection against allegations of infringement across the entire range of possible uses of that mark. In this case, even though the Singer's Mark had not acquired a reputation in relation to clothes as at the Priority Date, the Designer's Mark engaged s 60 because it was commonly accepted at the relevant time that persons whose trade marks had the reputation enjoyed by the Singer's Mark would sell clothing of a particular kind. That conclusion does not necessarily apply to other categories of goods.’
If you have not audited your trade mark portfolio recently, the High Court’s decision is a timely reminder to do so. Having a robust trade mark portfolio in place with standard and defensive coverage has the additional advantage of automatically blocking third-party filings for a deceptively similar mark or relevantly similar goods or services, without the need to oppose.
The outcome on the specific facts does not mean that famous trade marks are up for grabs as long as they are registered in respect of different goods or services or not registered at all.
During examination:
- Trade mark applications for famous names will often encounter an objection under section 43 of the Act, which can only be resolved with consent of the individual (this often occurs when a celebrity files for their name via a corporate vehicle).
- A famous name might also encounter an objection for being insufficiently distinctive under section 41 of the Act, as has recently occurred in respect of a trade mark application for GEORGE ORWELL in the EU.
As regards grounds of oppositions, whether a brand’s reputation precludes the registration of trade marks in respect of different goods or services under section 60 will always be highly fact-specific. The bad faith ground under section 62A could additionally come into play.
Background
The High Court’s decision stems from the enforcement of the registered trade mark KATIE PERRY by the Designer against the Singer. The Singer cross-claimed for cancellation of the Designer’s trade mark registration on various grounds.
At first instance, the primary judge dismissed the cross-claim. However, the Full Court of the Federal Court of Australia upheld an appeal, finding that the registration should be cancelled under both s 88(2)(a) and s 88(2)(c). We have previously summarised the Full Federal Court’s decision here.
Under s 88(2)(a), a court may cancel a trade mark registration based on any ground on which the registration could have been opposed. Relevantly, one such opposition ground is set out in s 60:
‘The registration of a trade mark in respect of particular goods or services may be opposed on the ground that:
- another trade mark had, before the priority date for the registration of the first‑mentioned trade mark in respect of those goods or services, acquired a reputation in Australia; and
- because of the reputation of that other trade mark, the use of the first‑mentioned trade mark would be likely to deceive or cause confusion.’
Additionally, under s 88(2)(c), a court may cancel a trade mark registration if, ‘because of the circumstances applying at the time when the application for rectification is filed, the use of the trade mark is likely to deceive or cause confusion’.
Under s 89(1), where one of these cancellation grounds is made out, the court may exercise its discretion not to cancel the registration ‘if the registered owner of the trade mark satisfies the court that the ground relied on by the applicant has not arisen through any act or fault of the registered owner’.
The majority judgments
Justice Jagot
At the outset, Jagot J addressed several matters of law raised by the appeal in relation to the application of ss 60, 88 and 89. Her Honour explained:
- Under s 60, a trade mark can only acquire a reputation in respect of particular goods or services.
- However, the reputation of a trade mark in respect of particular goods or services may mean that the use of an applied-for trade mark in respect of different goods or services would be likely to deceive or cause confusion for the purposes of s 60.
- Under s 88(2)(c), the question is whether deception or confusion is likely to arise from the notional use of the registered trade mark (that is, any normal and fair use within the scope of the registration). It does not call for consideration of the registered owner’s actual use of the trade mark.
- While s 88(1) uses the word ‘may’, it does not give the Court a discretion not to cancel a registration where a cancellation ground is made out. The only such discretion is in s 89.
- The ‘act or fault’ of the registered owner referred to in s 89 cannot be the mere act of applying to register the trade mark.
Her Honour then decided the appeal as follows.
The cancellation ground in s 88(2)(a) (with s 60) was not established. Before the priority date, the Singer had acquired a reputation in the trade mark KATY PERRY in respect of entertainment and recording discs. That reputation did not extend to clothes. And while it was possible that an ordinary person encountering the use of ‘KATIE PERRY’ in respect of clothes might momentarily wonder if it was somehow connected to the ‘KATY PERRY’ trade mark, in the absence of any apparent connection with the Singer, the danger of such confusion had not been proved to be real or tangible.
The cancellation ground in s 88(2)(c) was also not established. Her Honour disagreed with the Full Court’s finding that the primary judge placed undue weight on the lack of actual confusion:
‘[T]en years is a long time for two deceptively similar trade marks to be on the market in Australia with no evidence of any confusion, despite the attempts by [the Singer] to find such evidence by way of discovery. That period was sufficiently long to support an inference from the absence of evidence of confusion that in fact it was not likely that people had been or would be deceived or confused.’
In reaching these conclusions, her Honour gave primacy to the evaluation of the primary judge who had the benefit of hearing the entirety of the matter.
It was unnecessary to consider the exercise of the discretion in s 89(1). However, her Honour agreed with the Full Court that the relevant ‘act or fault’ of the Designer under s 89(1) would not have been merely applying to register her trade mark – it would have been doing so knowing about the Singer, her trade mark and the common practice of pop stars using their trade marks in connection with clothing.
Justice Steward
Justice Steward in general agreed with Jagot J. Where Steward J departed was in respect of section 88(2)(c). His judgment commenced by referring to four factual matters that he viewed as not having been given sufficient weight by the Full Federal Court:
‘First, much, but not all, of the merchandise sold in Australia was "Katy Perry" branded clothing, which was an infringement of the appellant's mark. Second, that infringing conduct was consistently pursued over a period spanning multiple years. Third, the infringing conduct occurred with knowledge, and in the face, of the appellant's mark, in circumstances where, in September 2009, the American artist had voluntarily amended her June 2009 application to register the "Katy Perry" mark in Australia to exclude goods in class 25. All of this conduct was very much deliberate. Fourth, this conclusion is not avoided because Bravado, under licence from Kitty Purry, sold the merchandise in question.’
Against that background, the concern was not to reward the efforts of an ‘assiduous infringer’. His Honour held that section 88(2)(c) should be interpreted in light of the maxim ‘“nullus commodum capere potest de injuria sua propria” – no man can take advantage of his own wrong’, in line with a number of High Court decisions decided under section 28(a) of the Trade Marks Act 1955 (Cth) (‘a mark – (a) the use of which would be likely to deceive or cause confusion; … shall not be registered as a trade mark’), namely Riv-Oland Marble Co (Vic) Pty Ltd v Settef SpA (1988) 19 FCR 569 and New South Wales Dairy Corp v Murray Goulburn Co-op Co Ltd (1990) 171 CLR 363.
His Honour held that whilst the literal terms of section 88(2)(c) refer to ‘circumstances applying at the time when the application for rectification is filed’ which would bring in the relevance of the efforts of ‘an assiduous infringer’, the provision should be ‘construed in a restricted or “strained” way so that it does not include such deception or confusion which has occurred because of the infringing conduct of an applicant for rectification’.
This approach could be reconciled with discretion in section 89(1), as that provision looks at the conduct of the proprietor of the trade mark, ‘rather than that of another person such as an assiduous infringer’. Section 89(1) might capture ‘changes in the market; changes in the products being sold; or because of the conduct of the proprietor of the mark, including how the mark has been or will be used’. Delay in pursuing infringement might also be relevant, albeit here there was good reason for such delay as the Designer ‘simply did not have the financial capacity’ earlier to pursue the American artist (that step was taken after the financial capacity was obtained via litigation funding).
Justice Gleeson
Justice Gleeson also agreed with Jagot J in the result.
In assessing s 60, her Honour explained that one relevant factor is ‘any common practice concerning the use of certain kinds of trade marks, to the extent that such practice is capable of demonstrating likely confusion on account of the identified reputation of the earlier mark’.
For both ss 60 and 88(2)(c), her Honour considered it relevant that an ordinary customer for clothes would have been aware that a pop star might seek to enhance the reputation of their trade mark by merchandising clothes or seek to build their brand by launching a clothing label.
Regardless, her Honour considered the likelihood of confusion remote. Her Honour explained that consumers:
‘…would probably expect clothes associated with [the Singer’s mark] to bear some association with Katy Perry as a pop star apart from the bare name “Katie Perry”, such as by displaying details of a concert tour by Katy Perry, or the name of a song or album released by Katy Perry. They would dismiss the possibility of a shared trade source where the clothes were not sold in a context that identified an association between Katy Perry and the clothes, such as at a musical performance by Katy Perry or on a website clearly associated with Katy Perry.’
The dissenting judgment
The judgment begins with the joint dissent of Gordon ACJ and Beech-Jones J. Their Honours agreed with Jagot J’s construction of the relevant provisions of the Act but differed on the application of those provisions to the facts at hand. (The dissent provides a handy summary of the facts and background from paragraphs 23 to 40 and of the first instance and Full Court decisions from paragraphs 41 to 50.)
Section 60
Reputation
The requirement that reputation (‘recognition of [the mark] by the public generally’) in s 60 be in respect of particular goods or services is repeated from the High Court’s earlier decision in Self Care v Allergan (2023) 277 CLR 186, as is the requirement that the resultant confusion in section 60(b) stem from reputation in the prior mark rather than reputation alone. The comparison is between the actual reputation in the prior mark as at the priority date of the opposed application, and the notional normal and fair use of the opposed mark (a hypothetical issue taking the use that could be made by the registrant, rather than the actual use it has made).
Gordon ACJ and Beech-Jones J accepted in light of unchallenged findings of the primary judge that recognition of the name ‘Katy Perry’ in Australia ‘was of necessity evidence of a reputation in that name as a mark’. However, their Honours disagreed with the Full Court’s reliance on the ‘“common practice” of pop stars to sell merchandise including clothing at concerts and to launch their own clothing labels’. This did not mean that the KATY PERRY mark had acquired a reputation in Australia in clothing before the priority date. The only reputation was in relation to music and entertainment. No merchandise had been sold or offered for sale at the priority date in Australia, even if this was not determinative (‘it is not necessary for the prior mark to have acquired a reputation in respect of the same goods or services in respect of which the opposed mark is proposed to be registered, although whether it has done so will be relevant to the likelihood of deception or confusion’).
Would use of the Designer’s Mark be likely to cause confusion due to the reputation in KATY PERRY at the priority date?
Gordon ACJ and Beech-Jones J summarised the key issue (‘the hypothetical question is whether a reasonable number of people may be caused to wonder if a trade connection exists between the Designer's Mark based on its notional normal and fair use on clothing and the Singer's Mark’) before concluding as follows, in line with the outcome before the Full Federal Court:
‘The notional normal and fair use of the Designer's Mark across all legitimate uses to which the Designer's Mark might be put by [the Designer] as at the Priority Date would extend to include the use of the Designer's Mark on labels of clothes as well as on the front of a t-shirt. The question whether the Designer's Mark is likely to deceive or cause confusion depends on whether as at the Priority Date it could be concluded that a reasonable number of people may be caused to wonder if a trade connection exists between the Designer's Mark based on its use on that clothing and the Singer's Mark. Or, put in different terms, as at the Priority Date would a reasonable number of people wonder whether a t-shirt bearing the Designer's Mark was connected with the Singer's Mark given the reputation that the Singer's Mark had in Australia at that time? The answer is "yes".’
This was premised on:
- The similarity between the marks. Whilst reputation is a threshold condition, the similarity between the marks can ‘increase or decrease the likelihood of deception or confusion by reason of the prior mark’s reputation’. Here, as with Jagot J, the dissent allows that ‘the strength of the reputation of a prior mark may mean that consumers notice small differences between trade marks so that, “giving the matter a moment’s reflection”, they “would readily conclude” that the trade marks were not related’. On the facts, the respective marks were deceptively similar (aurally identical) and the difference in spelling was not ‘striking’.
- The common practice of pop stars selling merchandise including clothing. The question was not whether the Singer’s reputation in entertainment and music was related to or extends to clothing. It was whether a consumer aware of the Singer’s reputation at the priority date confronted with a garment with the Designer’s Mark would be likely to be confused as to its source. That risk arose from the common practice. The combination between the reputation and that common practice meant that a notional buyer of clothing would be caused to wonder (‘It may be that a savvy consumer might have reflected on, for example, the absence of an image of [the Singer] or details of a recent tour date on a t-shirt, or the fact that the t-shirt was not being sold at a concert, and then concluded that there was no trade connection between the marks. That does not preclude the conclusion that there was a real, tangible danger that a notional buyer would be deceived or confused because of the reputation of the Singer's Mark’).
- The Full Court correctly finding that undue weight was given to the lack of instances of actual confusion by the primary judge. The inquiry is directed towards notional use. This would include prominent use of KATIE PERRY ‘on the front of a hoodie, t-shirt or tank top’. The lack of actual confusion could be explained by limited use of the Designer’s Mark.
- Finally, reliance on use of the Singer’s Mark ‘on clothes that also bore images of [the Singer]’ suggested that the primary judge was looking beyond normal and fair use.
Section 88(2)(c)
On section 88(2)(c), the threshold question was whether it should be assessed by reference to the actual use by the Designer or notional use. The relevant use ‘is necessarily the notional normal and fair use’, and the inquiry is not confined to actual use. This ‘properly reflects that the proprietor enjoys a statutory monopoly across the full range of goods or services to which the registration extends’ and would avoid permitting ‘the registered owner of a trade mark to resist rectification on the basis of its narrow use of the trade mark with a broad specification, even while there would be a real, tangible danger of deception or confusion if the mark was used beyond that narrow use’.
As at 20 December 2019, would that notional normal and fair use lead to a likelihood of deception or confusion? The considerations differ from section 60 (section 88(2)(c) is broader because the cause does not need to be the reputation in Australia of another mark), but ‘the applicable principles are analogous’. Here, the dissenting Justices took the view that the likelihood of confusion was much greater as at 19 December 2019, when the application for rectification was filed, than at the priority date, being the date at which the section 60 ground was decided. The Singer’s Mark had by that time been used on merchandise and by brand extension. There was no evidence ‘that at that time the Singer’s Mark was so notorious or well known that consumers would have been aware of the correct spelling of the Singer’s Mark’. The similarities between the marks tended to enhance rather than reduce the risk of confusion. The lack of any actual confusion did not detract, given that the use of the Designer’s Mark had achieved ‘little public presence’.
Section 89 – the discretion not to grant rectification
The onus of this provision was on the Designer to show that the grounds for cancellation did not arise through her ‘act or fault’. The broad term ‘act’ distinguished the concept from ‘blameworthy conduct’. Having said that, the collocation of act and fault meant that the registered owner’s state of mind was not irrelevant. Due to the use of ‘arisen by’, the act or fault need not be the sole cause of the ground for cancellation: ‘it is sufficient if the relevant act or fault caused or substantially contributed to the ground relied on’.
Taking the two points in time:
- Priority date (29 September 2008): The Full Court was correct to find that the section 60 ground arose through ‘[the Designer’s] act of filing the trade mark application for the Designer's Mark with knowledge of [the Singer], her reputation and her mark, and in circumstances where [the Designer] knew that popular music stars lend their names to be used in connection with clothes and sell merchandise at their concerts and accepted such conduct as a common practice’.
- Date of application for rectification (20 December 2019): In addition to the factors as at the filing date, the Designer ‘encouraged the association between herself and [the Singer] in the minds of consumers’ and failed to commence infringement proceedings for over ten years (‘during which time the reputation of the Singer’s Mark grew and the likelihood of confusion increased’).
The discretion was not enlivened.
