Taylor Swift's Trademark Filings Point to a New Legal Strategy Against AI Deepfakes
Taylor Swift's brand management company filed trademark applications in April 2024 covering short audio clips of her voice and her visual likeness. The move signals a shift in how the entertainment industry is fighting back against AI-generated impersonations.
Most AI-related lawsuits have focused on copyright - whether companies used copyrighted material to train their systems or whether AI outputs copied protected work too closely. The New York Times sued OpenAI and Microsoft over journalism used in training. Authors, publishers, and music labels have filed similar suits.
But copyright has a blind spot. It protects creative works, not identity. Copyright doesn't give Swift a general right to control anything that sounds like her or looks like her.
Why trademark matters for AI voices and faces
Trademark law protects names, images, sounds, and other markers that help consumers identify who stands behind a product or service. The MGM lion roar is a trademark. So is the Nike swoosh.
When an AI-generated voice imitates Swift to sell perfume or cryptocurrency, the real problem isn't copyright infringement. It's that listeners may think Swift approved the product. That's a trademark problem - one that involves deception about endorsement.
Swift's filings appear aimed at that danger. They target fake endorsements, fake appearances, and false signals of approval.
The publicity rights angle
Swift's concerns also overlap with publicity rights, which protect against unauthorized commercial use of a person's identity. A company using a celebrity's face in an advertisement without permission to suggest endorsement is a classic publicity rights violation.
AI's ability to clone voices and images makes this especially relevant. But publicity rights in the United States are governed by state law, and rules vary widely. That patchwork inspired the bipartisan NO FAKES Act, introduced in 2025, which would create a national standard prohibiting unauthorized AI-generated replicas of a person's voice or likeness. The bill is still in early stages at the Senate Judiciary Committee.
What remains untested
Courts have already affirmed that sounds can function as trademarks. Actor Matthew McConaughey trademarked "alright alright alright," his line from "Dazed and Confused," for similar protection.
But it remains unclear whether trademark law can police AI-generated replicas when the issue is manufactured endorsement rather than counterfeiting. A person's voice or likeness is not automatically a trademark - it must be used to help consumers identify who is behind a product or service.
Federal law also protects certain uses of a celebrity's image in parody, criticism, commentary, and news reporting. Not every imitation is deception.
Courts will have to draw distinctions case by case. A fake ad suggesting Swift endorsed a product differs from a parody commenting on celebrity culture. A scam using her voice differs from a news story about deepfakes.
The real problem Swift's filings address
AI has allowed fake endorsements to look and sound real enough to spread before correction is possible. Major AI copyright cases will continue focusing on copied works. But when AI manufactures identity, endorsement, or trust, copyright alone is insufficient.
Swift's filings suggest that AI law will increasingly focus not only on protecting the work of musicians, writers, journalists, and artists, but also on protecting the signals that tell audiences who is really speaking.
For creatives, this matters. As generative AI and voice technology advance, understanding the legal tools available - trademark, copyright, and publicity rights - becomes essential to protecting your identity and work.
Your membership also unlocks:
