Last month, global popstar Taylor Swift filed a trade mark application at the EUIPO and UKIPO for an audio recording of her voice, saying the phrase “Hey, it’s Taylor Swift”. These follow similar filings before the US and Australian intellectual property offices for recordings of the phrases “Hey, it’s Taylor Swift” and “Hey, it’s Taylor”.
Swift’s filing before the EUIPO office comes as generative AI makes imitating a celebrity’s voice or image easier than ever before. Swift joins a growing number of celebrities seeking greater control over the use of their image or voice in media, including Hollywood actor Matthew McConaughey and Italian Prime Minister Giorgia Meloni, both of whom have sought trade mark protection for specific audio and/or visual elements associated with them against the backdrop of AI’s increasing potential for imitation.
These filings raise the question of whether soundbites should be registrable and, if so, whether they can be enforced against AI deepfakes? Trade mark infringement generally requires the allegedly infringing sign to be used in the course of trade in relation to relevant goods or services and to satisfy one of the applicable infringement tests, which may include identity, a likelihood of confusion or, for marks with a reputation, unfair advantage or detriment. Taylor Swift has been the subject of numerous AI deepfake videos online in recent years, but is a deepfake video “use in the course of trade”? Furthermore, the scope of protection afforded by a sound mark covering the specific recording “Hey, it’s Taylor Swift” would not confer an abstract right over Swift’s voice and in practice provides a narrow scope of protection. Trade marks may therefore be a further tool to protect against unauthorised digital replicas and deepfakes, but the extent of their practical utility in this context remains to be tested.
Read more from our award-winning trade mark team here.
