Taylor Swift’s management company filed three federal trademark applications in late April 2026, seeking legal protection over her spoken voice and a specific onstage look. The filings, submitted to the U.S. Patent and Trademark Office by TAS Rights Management, mark one of the most aggressive moves yet by a major artist to lock down personal identity in an era when AI-generated clones can replicate a celebrity’s face, body, and vocal patterns in seconds.
The applications land more than two years after explicit AI-generated images of Swift spread across X (formerly Twitter) in January 2024, an episode that drew condemnation from the White House and fueled bipartisan calls for federal legislation. While Swift’s team has not publicly linked the new filings to that incident or any other specific deepfake, the timing and scope suggest a deliberate effort to arm her legal team with tools that go beyond copyright and state publicity laws.
What the filings actually cover
Two of the three applications are classified as sound marks. One covers the spoken phrase “Hey, it’s Taylor Swift.” The other covers the shorter greeting “Hey, it’s Taylor.” Sound marks are a recognized but uncommon corner of trademark law. The USPTO’s principal register contains roughly 250 active sound marks, most of them corporate audio logos: the NBC chimes, the MGM lion’s roar, the Intel bong. Registering a celebrity’s spoken greeting would push the category into new territory.
The third application is a visual mark describing a particular stage persona: Swift holding a pink guitar, wearing an iridescent bodysuit and silver boots. Visual marks of this kind function like a logo for a live performance identity. If granted, the registration would give Swift’s team a federal basis to challenge unauthorized reproductions of that concert image, whether created by AI, printed on bootleg merchandise, or used in advertising without permission.
All three applications are trackable through the USPTO’s public Trademark Search system and its Trademark Status and Document Retrieval portal. The specific serial numbers assigned to the filings can be located by searching for “TAS Rights Management” as the applicant in either tool. As of early May 2026, the filings are in the initial examination stage. No examiner decisions, office actions, or opposition proceedings have appeared in the public docket.
Why trademark law, and why now
Swift already controls her master recordings and songwriting copyrights, a hard-won position after her public battle with her former label. But copyright protects creative works, not a person’s identity. Right-of-publicity laws, which do protect name and likeness, vary wildly from state to state and offer no single federal enforcement mechanism. Trademark law fills a different gap: it protects consumers from being misled about the source of goods and services.
That distinction matters in the AI context. If a company uses a cloned version of Swift’s voice to say “Hey, it’s Taylor Swift” in a product ad, the argument for trademark infringement is straightforward: consumers could reasonably believe Swift endorsed the product. A registered sound mark would give her legal team a faster, more direct path to enforcement than piecing together claims under copyright or a patchwork of state laws.
The strategy also reflects a gap in federal legislation. The NO FAKES Act, introduced in both chambers of Congress in 2024 and reintroduced in 2025, would create a dedicated federal right protecting individuals from unauthorized AI replicas of their voice and likeness. But the bill has not reached a floor vote in either chamber. Without a purpose-built statute, performers are left repurposing older legal frameworks to address a problem those laws were never designed to solve.
“Trademark is not a perfect fit, but it’s a tool that already exists and already has federal teeth,” said Jennifer Rothman, a professor at the University of Pennsylvania Carey Law School who specializes in intellectual property and the right of publicity. Legal observers note that a successful registration would let Swift’s team send takedown notices grounded in federal trademark law, adding leverage alongside the copyright and defamation claims that already form the backbone of most platform enforcement requests.
The legal hurdles ahead
Sound marks face a higher bar than standard word or logo marks. The applicant must show that the sound has acquired “secondary meaning,” meaning consumers associate it with a specific commercial source rather than treating it as ordinary speech. The NBC chimes passed that test because decades of broadcasting cemented the three-note sequence as a network identifier. Whether the phrases “Hey, it’s Taylor Swift” and “Hey, it’s Taylor” carry the same kind of source-identifying power is an open question that USPTO examiners will have to evaluate.
The visual mark presents its own challenges. Trademark examiners will assess whether the described combination of elements, pink guitar, iridescent bodysuit, silver boots, is distinctive enough to function as a source identifier rather than simply a costume. Concert imagery is inherently expressive, and the line between a protectable brand element and a performer’s artistic choice is not always clear.
If the applications are refused, the refusals themselves could set informal precedent, signaling to other artists and their lawyers how skeptical the USPTO is about extending trademark protection to personal identity markers. If the applications succeed, they could open the door to a wave of similar filings from musicians, actors, and athletes looking to preempt AI-driven impersonation.
What this means for fans and platforms
For everyday listeners and concertgoers, the practical impact is narrow. Trademark registration does not prevent anyone from saying Swift’s name, quoting her lyrics under fair use, or posting concert photos taken from the audience. It targets commercial misuse: a company deploying a cloned voice to imply Swift’s endorsement, or a vendor selling merchandise that replicates her trademarked stage look.
Platforms and intermediaries may feel the effects more directly. Streaming services, ad networks, and social media companies already process takedown requests rooted in copyright. If Swift secures these registrations, her team could layer trademark claims onto those notices, increasing pressure on platforms to remove AI-generated content that simulates her voice or concert persona in a commercial context. That added legal weight could, over time, influence how platforms draft their synthetic media and impersonation policies.
A test case the entertainment industry is watching closely
Other artists and their legal teams are paying close attention. Swift’s filings represent one of the first attempts by a major performer to use trademark law as a frontline defense against AI cloning. The outcome, whether the marks are granted, refused, or narrowed during examination, will shape how the entertainment industry thinks about identity protection for years to come.
The examination process typically takes several months, and any resulting registrations could face opposition from third parties during a public comment window. Interested readers can monitor the USPTO docket directly through the agency’s trademarks portal for status updates as the applications move forward. Until examiners weigh in, and until courts have a chance to interpret any registrations that result, the Swift filings will remain one of the sharpest early tests of whether traditional intellectual property law can stretch far enough to meet the challenges posed by generative AI.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
