The Backstreet Boys trademark their voice amid the rise of AI deepfakes

By Digital Music News · June 26, 2026.

The Backstreet Boys have officially filed a sound mark application with the United States Patent and Trademark Office (USPTO) through their company BSB Entertainment. The application, spotted by the specialist firm Gerben IP, seeks protection for the spoken words 'HI, WE'RE THE BACKSTREET BOYS', accompanied by audio specimens taken from concert announcements where the group utters that phrase —along with its variant 'Hey, we're the Backstreet Boys'—. The scope of the mark would cover 'audio and video recording services', including live performances, prerecorded recordings and multimedia websites.

With this move, the band —which has just marked 33 years of career— joins a growing trend among top-tier artists seeking to legally shield their vocal identity against the advance of generative artificial intelligence. Taylor Swift has already applied for sound marks for the phrases 'Hey, it's Taylor' and 'Hey, it's Taylor Swift', while Lionel Richie filed for protection of spoken fragments of his most iconic songs, such as 'Easy like Sunday morning' and 'All night long'. Actor Matthew McConaughey, for his part, has already successfully obtained multiple registered voice marks, making his case one of the clearest precedents in the sector.

Sound marks are not a new legal mechanism, but they do represent, in the words of Gerben IP, 'a lesser-known category of trademark protection'. The most famous example in the entertainment world is Netflix's 'tu-dum' sound, registered as a trademark. Outside the cultural industry, companies such as Lotus Bakeries have gone so far as to register advertising phrases spoken in a specific vocal tone —in their case, 'You gotta love Biscoff', described as 'confident, warm, slightly smoky and with a drawn-out pronunciation of the word love'—, which demonstrates that the mechanism has applicability across multiple sectors.

The stated motivation behind these applications is twofold: on one hand, to provide artists with an additional legal tool to fight AI-generated voice deepfakes in court; on the other, to open the door to proactive licensing agreements with artificial intelligence platforms. On this second front, the article notes that companies such as ElevenLabs and Klay Vision have already opted for licensing routes to avoid costly litigation.

However, the real effectiveness of these sound marks against generative AI remains a legal unknown. Gerben IP itself acknowledges that it is unclear whether a mark registered for a specific phrase can be used to pursue any kind of sound content that imitates an artist's voice without exactly reproducing that phrase. In other words: protecting 'Hi, we're the Backstreet Boys' does not automatically imply protection against any deepfake audio that simulates the group's voices in other contexts.

From the perspective of agentic AI and audio generation models, this move has significant implications. Voice synthesis and voice cloning tools —already integrated into agentic workflows for music production, automated dubbing or conversational assistants— operate in a legal ecosystem that is still under construction. If courts validate these marks as an effective instrument against the unauthorized use of cloned voices, developers of AI agents that generate audio will have to incorporate additional layers of checking and filtering to avoid infringements, which would add technical and legal friction to their pipelines.

In parallel, the sector could evolve toward a structured licensing model: artists register their voices as trademarks, and AI platforms negotiate usage rights the same way sync or master rights are negotiated today in the traditional music industry. Companies like ElevenLabs, which according to the source has opted for licensing approaches, would be getting ahead of that model. If the movement to register vocal trademarks becomes widespread —as the rapid adoption by figures across different genres and generations seems to suggest—, the market for synthetic voice licensing could become a significant new revenue stream for artists and a new structural cost for generative AI platforms.

For now, the process is in its early stages. The applications from the Backstreet Boys, Swift and Richie must still pass USPTO examination before obtaining definitive registration. The coming months —and any legal disputes that may arise— will provide the first real jurisprudence on how far the protection of these marks extends in the specific context of AI deepfakes.