The Backstreet Boys have sold more than 150 million records and spent three decades as one of the most recognizable acts in popular music. Now they are trying to do something that has never been done quite this way before: trademark the sound of their own voices.
According to reporting from Billboard, the group is seeking federal trademark protection for their individual vocal signatures, joining Taylor Swift and other artists racing to establish legal footholds against AI-generated voice clones. The move is creative, legally ambitious and ultimately a symptom of a much larger problem: the law has not kept pace with what artificial intelligence can now do to a person’s identity. Indeed, modern trademark law does not protect a voice in gross; it protects specific sounds used as source identifiers.
What a Sound Mark Application Actually Requires
Most people understand trademarks as brand names and logos — the kind of visual, word-based identifiers that show up on packaging and storefronts. A sound mark works on the same legal theory, but the application process is materially harder.
Under Section 1202.15 of the USPTO’s Trademark Manual of Examining Procedure (TMEP), a sound mark “identifies and distinguishes a product or service through audio rather than visual means.” To function as a protectable source identifier, the sound must “assume a definitive shape or arrangement” and “create in the hearer’s mind an association of the sound” with a specific good or service. The TMEP recognizes registrable sound marks as including a series of tones or musical notes, and wording accompanied by music, but draws a firm line at “commonplace” sounds that consumers encounter routinely and therefore do not associate with any particular source.
The application itself requires more than a standard filing. Applicants must submit an audio file of the precise sound mark in .wav or .mp3 format, along with a written description detailed enough for an examiner to evaluate. Specimens showing actual use in commerce, meaning the sound being used to identify goods or services in the marketplace, not just performed as music, are also required.
That last requirement is where voice mark applications get complicated. For a well-known performer, proving that consumers recognize the voice is not particularly difficult. Proving that they associate it with a specific commercial source, the way a consumer hearing NBC’s three-note chime thinks “NBC” rather than simply “familiar sound”, is a higher bar. A voice that has spent thirty years as a musical act needs to demonstrate it also functions as a brand identifier in a commercial sense, not merely an artistic one.
Beyond recognition, the USPTO typically requires applicants to show acquired distinctiveness, also known as secondary meaning. Under Section 2(f) of the Lanham Act, this requires evidence that the mark has become distinctive of the applicant’s goods or services through substantially exclusive and continuous use in commerce, a threshold that can be met by five or more years of exclusive use, consumer surveys, media coverage, or other corroborating evidence. For the Backstreet Boys, whose voices have been commercially ubiquitous for three decades, that evidentiary record is unusually strong. But the inquiry does not stop at recognition: the USPTO’s examining attorney must also be persuaded that the sound is “arbitrary, unique or distinctive”, and that it would not be treated as merely a generic or functional sound in the relevant industry context.
The practical contrast with a word mark is significant. When a company applies to register a brand name, distinctiveness can often be established on the face of the mark itself, an invented word like “Kodak” is inherently distinctive and needs no secondary meaning evidence at all. A sound mark, by contrast, is almost never inherently distinctive. It will nearly always require a showing of acquired distinctiveness, which means the application process begins with a heavier evidentiary burden and a higher probability of an Office Action from the examiner before registration proceeds.
Billboard also notes what a successful registration would and would not provide: even with a granted trademark, it remains far from certain that the registration would give the Backstreet Boys real legal power to stop someone from using their voice for different words. Trademark infringement requires a likelihood of consumer confusion about the source of goods or services. An AI clone singing unauthorized lyrics may cause real harm, but whether that harm constitutes the kind of commercial source confusion trademark law was designed to address is a question courts have not yet resolved in this context.
What the Current Law Does (and Doesn’t) Cover
Absent a specific trademark registration, performers have historically relied on three overlapping bodies of law to protect their vocal identity.
Right of publicity laws protect an individual’s name, likeness and, in many states, voice from commercial exploitation without consent. These protections vary significantly by jurisdiction. Some states offer robust, codified rights; others recognize only a narrow common-law tort. There is no federal right of publicity statute, which means enforcement is a patchwork exercise and actors in permissive states face real exposure.
The Lanham Act’s false endorsement doctrine offers a related federal hook. Courts have recognized that using a celebrity’s voice in a way that implies endorsement can constitute unfair competition. The Ninth Circuit’s decisions in cases involving Bette Midler and Tom Waits established early precedents here, but those cases involved human performers being imitated, not algorithmically reconstructed replicas trained on years of recorded material.
Copyright law protects the recordings and compositions themselves but does not protect a person’s voice as a standalone creative work. The sound of someone speaking or singing is not, by itself, a copyrightable work.
The result is a gap because each doctrine protects only a slice of the problem, and AI exploits the seams between them.
Congress Is Moving — Faster Than Usual
The legal picture may shift materially in the near term. The NO FAKES Act of 2026, the Nurture Originals, Foster Art, and Keep Entertainment Safe Act, passed the Senate Judiciary Committee unanimously on June 18. The bill would create the first federal intellectual property right in an individual’s voice and visual likeness, require platforms to remove unauthorized deepfakes, and expose knowing violators to penalties of up to $750,000 per unauthorized work.
Notably, the coalition behind the bill is unusually broad. Supporters include SAG-AFTRA, the RIAA, Google, YouTube, OpenAI, Disney and the AFL-CIO. The bill preserves First Amendment carveouts for parody, news, satire, commentary and research, and it extends protections postmortem, allowing estates to license or enforce rights for up to 70 years after death.
Whether the bill reaches the President’s desk this Congress remains uncertain, but its unanimous committee passage signals political will that has not been present before.
What Businesses and Brand Owners Should Be Doing Now
The Backstreet Boys’ trademark strategy reflects a broader lesson that applies well beyond the music industry: in the current environment, proactive IP protection matters. Waiting for a clear statutory framework before taking action is a risk.
For brand owners, performers and companies that hold valuable voice or likeness assets, whether individual talent, a proprietary AI-generated spokesperson or audio branding, several steps are worth considering right now.
First, audit what you have. If your business uses a distinctive voice in advertising, interactive products or AI tools, evaluate whether that voice may be eligible for sound mark protection and whether it is currently being monitored for unauthorized replication.
Second, review your talent agreements. Existing contracts likely predate the current AI environment. Whether your agreements address voice cloning rights, synthetic replica licensing and deepfake consent is a question worth answering before a dispute forces the issue.
Third, watch the NO FAKES Act closely. If it passes, the compliance obligations for platforms that host user-generated content will change significantly, and companies that have not thought through their exposure should do so now.
The Backstreet Boys are telling us something important: even the most recognizable voice in the world is not legally protected by default. In an era when any voice can be replicated at scale for the cost of a cloud subscription, that is not a comfortable place to be.
Contact
To discuss further, contact KJK Partner Kyle Stroup (KDS@kjk.com).