Choose Your Battles: Katy Perry Loses Trade Mark Fight to Australia’s “Katie Perry” Taylor v Killer Queen LLC [2026] HCA 5

Can Katy Perry and Katie Perry coexist on a clothing rack? Yes, and the High Court found a risk of confusion is far smaller than fame might suggest.

In the 3:2 decision, the High Court has allowed the appeal of Sydney fashion designer Katie Jane Taylor (Taylor), setting aside the Full Federal Court’s cancellation of the “Katie Perry” mark for clothing, remitting the unresolved aspects of the case back to the Full Court. 

The High Court was unanimous on the questions of statutory construction; it only found contention in how those provisions applied to the facts at hand. The lesson for practitioners is not that fame is irrelevant, but that fame is not substitute evidence of reputation. An applicant for rectification must still prove that, because of the reputation it has established, use of the challenged mark is likely to deceive or cause confusion.

Background

Taylor, born Katie Jane Perry, is a fashion designer who sold clothing, particularly luxury loungewear, under her own name since around 2007. Taylor registered “KATIE PERRY” as a business name in April 2007 and began selling her luxury-wear under the name in 2008.

On 29 September 2008, Taylor applied for the word mark in class 25 (clothing). That filing date is the priority date, pursuant to s 12 of the Trade Marks Act 1995 (Cth) (TMA). The mark was entered onto the Register on 21 July 2009.

On the other side of the world, Katheryn Hudson (Hudson) had performed as “Katy Perry” since roughly 2002. Hudson’s album One of the Boys had released in June 2008, and by mid-2008 her singles were charting in Australia. Taylor accepted that Katy Perry was an internationally famous pop star from about July 2008. Before the mark’s priority date, no clothing items branded “Katy Perry” had been distributed in Australia.

In June 2009 Ms Hudson applied to register “Katy Perry” in classes 9, 25 and 41 (recording, clothing and entertainment). After an adverse examination report found that the proposed mark closely resembled Taylor’s registration, Hudson amended the application to exclude class 25. The mark was registered in November 2011 in classes 9 and 41. Branded clothing continued to be sold under the name despite this filing.

Procedural History

On 24 October 2019, Taylor commenced infringement proceedings in the Federal Court. Justice Markovic found that Kitty Purry Inc, one of Hudson’s associated companies, had infringed Taylor’s mark and dismissed a cross-claim by Hudson and Killer Queen LLC seeking its cancellation under ss 88(2)(a) and 88(2)(c) of the TMA. On appeal, the Full Court held that both grounds for cancellation were established and ordered that Taylor’s mark be cancelled. Taylor then appealed to the High Court.

Issues Before the High Court

The appeal concerned the operation of two grounds for cancellation.

Ground One

• Section 88(2)(a) read with s 60, speaks to reputation at the priority date. Section 88(2)(a) allows cancellation on any ground on which registration could have been opposed. The relevant opposition ground was s 60, which requires proof that another trade mark had acquired some reputation in Australia before the priority date, and that because of said reputation, use of the challenged mark would be likely to deceive or cause confusion.

Ground Two

• In the alternative, s 88(2)(c) permits cancellation where, because of the circumstances applying when the rectification application was filed, use of the mark is likely to deceive or cause confusion. The assessment date was 20 December 2019, by which time Hudson’s fame and merchandising activities had grown considerably.

Had either ground been established, s 89(1) confers a discretion not to rectify, available only where the registered mark owner satisfies the court that any relevant ground has not arisen through any act or fault of the owner.

The High Court’s Findings

The High Court ultimately allowed Taylor’s appeal, restoring the primary judge’s finding, and set aside the Full Court’s cancellation. Critically, all dissentients agreed with Jagot J’s statutory construction; the matter only concerned the application.

a) Reputation Attaches to Goods, But Not Necessarily the Same Goods

Justice Jagot answered three questions of reputation. (1) trade marks cannot acquire reputation “at large”, or disconnected from goods or services; (2) reputation in one class of goods does not, of itself, amount to reputation in another; and (3) use may nevertheless be likely to deceive or cause confusion under s 60(b) because of a reputation acquired in different goods or services — a question of fact to be decided in each case.

This means a reputation in different goods or services can still ground an objection. Justice Gleeson held the text of s 60(a) imposes no requirement that the two sets of goods match, while whether the reputation is likely to deceive is a separate factual question under s 60(b).

At the September 2008 priority date, “Katy Perry” was famous in Australia for music and entertainment, but the evidence did not establish that this reputation made consumers likely to associate Taylor’s clothing with Hudson; accordingly, s 60 was not established.

b) Common Practice Cannot Manufacture Reputation

The Full Court had reasoned that the common practice of pop stars selling merchandise and launching clothing labels meant that reputation could not be confined to entertainment and music. Justices Jagot and Gleeson rejected that step. Evidence of general merchandising practices could not, without evidence, establish that the “Katy Perry” mark had acquired a reputation for clothing in Australia.

c) Confusion Was Not Established

Testing the notional normal and fair use of the mark against Hudson’s reputation, the majority found no real danger of confusion. A consumer in 2008 who knew of Katy Perry would expect clothing from that source to carry some signal of the connection; absent this, the danger of confusion was “possible but has not been proved … to be real or tangible”.

Although Hudson’s reputation had grown substantially by December 2019, ten years of concurrent use without evidence of actual confusion supported the finding that confusion 

Additional Observations

a) An “assiduous infringer”

Justice Steward held that s 88(2)(c) must be read down. A party should not be able to infringe someone else’s valid trade mark, create confusion through its own conduct, and then rely on that confusion to have the trade mark cancelled. The other majority judges did not adopt this view.

b) Application for Rectification

Because neither ground for cancellation was established, the s 89 discretion of the courts to rectify the Trade Marks Register never arose. [EF1] 

c) IP Australia’s Manual Update

Following the judgment, IP Australia updated its Trade Marks Manual in March 2026 so that Part 46.4 now records that a s 60 reputation “must be reputation as a trade mark … and must be in respect of particular goods and/or services”, citing this case.

Key Takeaways for Practitioners

• Prove reputation and confusion separately. Fame or industry practice alone will not establish reputation in particular goods or services. Evidence that celebrities commonly sell merchandise may support an argument about confusion, but it does not extend a mark’s reputation.
• Reputation need not concern the same goods. Reputation in another class of trade mark may support s 60 if the evidence establishes likely confusion.
• Long coexistence without actual confusion is persuasive. Keep records of complaints, misdirected enquiries and returns.
• Greater fame may reduce confusion. Consumers may more readily recognise that unrelated goods have no connection with the famous brand.