She's not waiting for the law to catch up. She's building it herself.
In an age when technology can conjure a person's face, voice, and convictions without their knowledge or consent, Taylor Swift has turned to an unlikely instrument of self-defense: the trademark. By filing three applications with the U.S. Patent & Trademark Office — two covering her spoken phrases, one her visual likeness — Swift is attempting to extend the reach of intellectual property law into territory that existing celebrity-protection statutes have struggled to hold. The move reflects a broader reckoning across the entertainment world, where the most recognizable human beings are discovering that their very identities have become raw material for systems they never authorized.
- Swift's likeness has been weaponized repeatedly — appearing in AI-generated pornographic images, unauthorized commercial contexts, and fake political endorsements shared by Donald Trump ahead of the 2024 election.
- Existing right-of-publicity laws vary too widely by state to offer reliable protection, leaving performers exposed to harms that outpace the legal frameworks meant to address them.
- By filing sound and visual trademarks through her company TAS Rights Management, Swift is borrowing a strategy pioneered by Matthew McConaughey's legal team to open new enforcement pathways against AI misuse.
- The applications — flagged by IP attorney Josh Gerben — are untested at scale, and neither Swift's representatives nor the filing attorney have commented publicly on the strategy.
- If the trademarks hold, unauthorized commercial use of her voice phrases or image — including by AI systems trained to replicate her — could become actionable in ways that go beyond current celebrity law.
- The entertainment industry is watching closely: a successful legal precedent here could reshape how courts and legislators define identity protection in the AI era.
Taylor Swift has spent years watching unauthorized versions of herself spread across the internet — deepfakes, pornographic images, and AI-generated political endorsements she never made. Now, through her company TAS Rights Management, she is fighting back with an unconventional legal tool: the trademark.
Swift recently filed three applications with the U.S. Patent & Trademark Office, a move first identified by intellectual-property attorney Josh Gerben. Two are sound trademarks covering the phrases "Hey, it's Taylor Swift" and "Hey, it's Taylor." The third is a visual trademark describing a precise image of Swift — pink guitar, iridescent bodysuit, silver boots, purple-lit stage. The filings are unusual because trademark law has not traditionally been used to protect a person's voice or appearance; that work has fallen to right-of-publicity claims, which vary significantly in strength from state to state.
The theory, apparently borrowed from actor Matthew McConaughey's legal approach, is that trademark protections could reach further than existing celebrity law — making unauthorized commercial use of her voice or image, including by AI systems trained to mimic her, legally actionable in new ways. The documented misuse of Swift's identity has been severe: Meta's AI tools used her image without permission, pornographic deepfakes circulated widely, and in the weeks before the 2024 presidential election, Donald Trump shared AI-generated images falsely depicting her as a campaign supporter.
Neither Swift's representatives nor the attorney listed on the filings have commented publicly. The strategy remains untested at scale. But the broader significance is hard to miss: Swift is among the most recognizable people on earth, which makes her a frequent target — yet the underlying vulnerability belongs to any performer with a distinctive voice or face. By attempting to build legal architecture through trademark filings rather than waiting for legislation to catch up, she is signaling that the most exposed individuals are no longer content to wait.
Taylor Swift has spent the last several years watching versions of herself proliferate across the internet — versions she never approved, never performed, never consented to. Now she's fighting back through the U.S. Patent & Trademark Office.
Swift's company, TAS Rights Management, recently filed three trademark applications with the USPTO, a move first spotted by intellectual-property attorney Josh Gerben of Gerben IP. Two of the applications are sound trademarks covering her voice: the phrases "Hey, it's Taylor Swift" and "Hey, it's Taylor." The third is a visual trademark describing a specific image — Swift holding a pink guitar with a black strap, dressed in a multi-colored iridescent bodysuit and silver boots, standing on a pink stage in front of a multi-colored microphone with purple lights behind her.
The filings are unusual. Trademark law has not traditionally been the tool of choice for protecting a person's voice or general appearance — that's the territory of right-of-publicity claims, which vary widely in strength from state to state. But the theory, which Swift appears to be borrowing from actor Matthew McConaughey's legal team, is that trademark protections could open up additional legal avenues against AI-generated content that exploits someone's identity without permission. Where right-of-publicity claims have limits, trademark infringement claims might reach further.
Gerben, who flagged the applications, described them as a direct response to the entertainment industry's mounting anxiety about artificial intelligence. Performers have watched AI systems learn to replicate their voices, faces, and mannerisms with increasing accuracy — and then watched that technology get used against them. For Swift, the threat has not been abstract.
Her likeness has been misappropriated in a string of documented incidents. Meta's AI chatbots have used her image without authorization. Pornographic deepfakes featuring her face circulated widely online. And in the weeks before the 2024 U.S. presidential election, Donald Trump shared AI-generated images falsely depicting Swift as a supporter of his campaign — images that had no basis in reality and that she had no hand in creating.
Representatives for Swift have not commented on the filings, and Rebecca Liebowitz, the partner at law firm Venable listed on the applications, has also stayed quiet. The legal strategy is still untested at scale, but the logic is clear enough: if a phrase or image can be registered as a trademark, then unauthorized commercial use of that phrase or image — including by an AI system trained to mimic the real thing — becomes actionable in ways that go beyond what existing celebrity-protection law typically allows.
The broader stakes here extend well past Swift's career. She is among the most recognizable people on the planet, which makes her a frequent target, but the underlying problem belongs to every performer with a distinctive voice or face. The entertainment industry has been lobbying for stronger federal protections against AI misuse, and high-profile legal maneuvers like this one tend to shape how courts and legislators think about what the law should cover.
Whether Swift's trademark applications will ultimately hold up — and whether they'll prove effective against the specific harms she's experienced — remains to be seen. But the filings signal something important: that the people most exposed to AI-generated identity theft are no longer waiting for the law to catch up on its own. They're trying to build the legal architecture themselves, one application at a time.
Notable Quotes
The trademark applications reflect growing concern among entertainment talent about AI stealing artists' ability to control their voice and likeness without consent.— Josh Gerben, intellectual-property attorney, Gerben IP
The Hearth Conversation Another angle on the story
Why trademarks? Why not just sue under existing right-of-publicity laws?
Right-of-publicity claims are inconsistent — strong in some states, weak in others, and not always well-suited to the speed at which AI content spreads. Trademark law offers a different set of remedies and a federal framework that's harder to dodge.
So this is essentially a workaround for a gap in the law?
More like a bridge. The law hasn't caught up to what AI can do, so her legal team is using an older tool in a new way — the same move McConaughey's lawyers tried first.
What does it mean to trademark a phrase like "Hey, it's Taylor"?
It means that phrase, used in a commercial context to identify her, becomes legally hers. If an AI product uses it to simulate her voice or sell something, that's potentially infringement — not just impersonation.
And the visual trademark — a very specific photograph. Why that level of detail?
Trademarks protect specific, identifiable marks, not general appearances. The more precisely you describe what you're protecting, the stronger the claim. That pink guitar, that bodysuit — it's a recognizable image tied to her brand.
Does this actually stop deepfakes from being made?
Not directly. It doesn't prevent creation, but it creates legal consequences for distribution and commercial use. That's where the real harm tends to happen anyway.
The pornographic deepfakes, the fake Trump endorsement — those feel like very different kinds of harm.
They are. One is sexual exploitation, the other is political manipulation. Both happened without her consent. The trademark strategy addresses the commercial and reputational dimensions, but it's not a complete answer to either.
What does it mean that she's doing this now, in 2026, rather than earlier?
It means the incidents accumulated to a point where waiting for legislation felt untenable. She's not the only one — she's just the one with the resources and visibility to move first in this particular direction.