Co-authored by Gowling WLG articling student Sarah Bonner-Proulx.
 

By now, Taylor Swift is almost as well known for her branding strategy as she is for her music. From carefully curated “eras” to extensive trademark portfolios protecting lyrics, album titles, and merchandise, Swift has long treated intellectual property as an important component of her celebrity brand. Her latest filings with the United States Patent and Trademark Office (“USPTO”), however, may represent something different entirely: a proactive attempt to protect her voice and likeness from artificial intelligence.

Through her company, TAS Rights Management, Swift has filed three trademark applications with the USPTO: one for a visual image associated with the Eras Tour, and two for short audio clips— “Hey it’s Taylor” and “Hey it’s Taylor Swift”—recorded during promotion for her 2025 studio album, The Life of a Show Girl (which is currently the subject of a trademark dispute).

More than a traditional trademark filing

Unlike traditional trademark filings for names or designs, or even nontraditional trademarks for motions, sounds, or texture, these applications combine conventional and unconventional elements, focusing on features that are more closely tied to personal identity itself. The visual filing covers a widely recognizable Eras Tour photograph featuring Swift with a pink guitar, iridescent bodysuit, and silver boots. The audio filings target short vocal phrases associated directly with Swift’s voice and persona.

The filings arrive against the backdrop of growing concern within the entertainment industry over AI-generated deepfakes, voice cloning, and unauthorized digital replicas. As generative AI tools become increasingly sophisticated, musicians, actors, athletes, and other public figures are facing new challenges in controlling how their likenesses and voices are reproduced online.

In that context, Swift’s filings appear strategically designed to expand the tools available to combat unauthorized AI-generated content.

Trademark law as an AI protection tool

Public figures already benefit from certain “right of publicity” protections, which generally prevent the unauthorized commercial use of a person’s name, image, likeness, or identity. However, those protections are often fragmented and jurisdiction-specific.

In the United States, publicity rights largely operate at the state level, meaning both the scope of protection and available remedies vary considerably across jurisdictions. In Canada, similar protection is derived from a mix of common law (e.g., the tort of misappropriation of personality) and statutory frameworks (e.g., federal and provincial privacy statutes). In both jurisdictions, enforcement is particularly challenging in fast-moving digital environments where infringing content spreads rapidly across online platforms.

Trademark law may offer a more structured enforcement mechanism.

A federal trademark registration in the US or Canada can provide nationwide enforceability, procedural advantages under the United States Lanham Act or the Canadian Trademarks Act, respectively, and access to remedies not always available under publicity rights or misappropriation of personality frameworks. Importantly, trademark law focuses on consumer confusion and source association, concepts that may become increasingly significant as AI-generated content becomes harder to distinguish from authentic material.

Whether these particular applications ultimately succeed remains to be seen. Trademark law was not originally designed to address synthetic voices or AI-generated replicas, and the USPTO and the Canadian Intellectual Property Office (“CIPO”) may scrutinize whether the filings function as indicators of source in the traditional trademark sense. Nevertheless, the applications themselves send an important signal.

The ripple effect

The entertainment industry is no longer waiting for legislatures to catch up with the rapid advances in artificial intelligence. Instead, this willingness by a public figure to leverage every available intellectual property tool to assert control over their voice, image, and digital identity may signal a broader shift. If successful, Swift's approach could become a template not only for musicians, but also for athletes, influencers, broadcasters, and other public-facing personalities seeking to protect themselves against AI misuse.

Despite her trademark application still pending approval, Swift has inspired other musicians to seek out their own sound-mark applications. In June 2026, Lionel Richie filed four voice and likeness trademarks of snippets of his famed songs, one of which was for the spoken phrase “Easy like Sunday morning.” Later that same month, the Backstreet Boys’ entertainment company submitted a sound-mark trademark application, like Swift’s, but this time for the spoken words “hi, we’re the Backstreet Boys.”

These developments suggest that Swift’s strategy is not an isolated move but rather marks the beginning of a broader wave of artists using trademark law as a shield against unauthorized AI reproductions.

Key takeaways

Taylor Swift's pending trademark filings may reflect the beginning of a broader shift in how public figures respond to the rise of generative AI. As deepfakes, voice cloning, and digital replicas become increasingly sophisticated, this case illustrates how entertainers may begin turning to existing intellectual property frameworks, including trademark law, to assert greater control over their voice, image, and identity.

Regardless of whether the applications are ultimately approved, the filings send a clear message: talent is no longer waiting for legislatures to catch up before taking proactive steps to protect their digital presence.

Whether you are seeking to run a trademark search or enforce your trademark, or have any other trademark needs, we can assist. Please feel free to contact one of the authors or a member of our Entertainment & Sports Sector group if you have any questions.