No Federal Voice Rights Yet, Where AI Protection Actually Stands for Artists

An AI-generated voice can sound convincingly like an artist, but there is not yet a U.S. federal law specifically establishing rights against AI voice clones. For now, artists are working with a patchwork: state laws, platform policies and the terms they negotiate in private agreements.

That gap matters. Legal protections are still developing while tools for creating and distributing synthetic audio are already in use. As the Recording Industry Association of America (RIAA) and trade publications including Music Business Worldwide, Complete Music Update and The Independent have reported, lawmakers and music-industry organizations are debating how to protect voices and likenesses, and how to license AI uses fairly.

What the NO FAKES Act would do, and where it stalled

The NO FAKES Act of 2026 (S. 4591) would create a federal intellectual property right allowing people to authorize the use of their voice or visual likeness in a digital replica. The bill addresses replicas that imitate a person, rather than ordinary uses of copyrighted sound recordings or compositions.

The measure cleared the Senate Judiciary Committee in June 2026. On September 30, Senator Marsha Blackburn sought to advance it through unanimous consent, but an objection from Senator Ted Cruz prevented that route from succeeding. This was not a recorded vote rejecting the bill, and it did not make the proposal law. It remains pending.

The RIAA supports the bill. Its public statements describe a federal right against unauthorized digital replicas, while reporting by Music Business Worldwide has covered the bill’s progress and the music industry’s interest in the issue. With the midterm elections scheduled for November 3, 2026, industry figures are looking to a possible lame-duck session. Reintroduction in 2027 is also considered likely.

Until Congress enacts a federal law, there is no U.S. federal statute establishing a specific right against AI voice clones. Other laws may apply in particular circumstances, but artists should not assume that a single federal rule currently covers every unauthorized synthetic voice.

A studio workstation displays abstract waveforms beside a mixing desk, suggesting synthetic voice replication

What protects artists today

State laws

Some states have laws addressing a person’s voice or likeness. Tennessee’s ELVIS Act, enacted in 2024, specifically added protections related to voice, including certain unauthorized uses of a person’s voice and tools used to create an identifiable voice replica.

State rules are not uniform. Their scope, exceptions and available remedies can differ, and a law that applies in one state does not automatically provide a nationwide solution. Artists should also distinguish a voice or likeness claim from copyright in a song or recording: they can involve different rights and different questions.

Platform takedown tools

Some platforms have policies or reporting processes for impersonation and manipulated media. For example, YouTube’s impersonation policy and AI-content guidance describe rules and processes relevant to certain synthetic or misleading content.

These tools can help an artist report a suspected clone or impersonation. But a platform policy is not the same as a law. The platform decides how its rules apply, and a takedown process may not settle who owns a right, whether a use was legally permitted or what compensation is due.

Your own contracts

A contract can set boundaries before a voice or likeness is used. Recording, producer, feature, sync and other agreements may address whether material can be used to train AI systems or generate synthetic performances. The wording matters: broad or unclear permissions can create disputes later.

Artists should read the scope, duration and payment terms carefully, including clauses about future technologies, sublicensing and reuse. For consequential agreements, advice from an entertainment attorney can help clarify the rights being granted and retained.

The difference between a right and a remedy

A right defines what a person may control or prohibit. A remedy describes what options may be available if that right is violated, such as seeking removal, bringing a claim or requesting compensation. Having a platform reporting tool does not necessarily establish a legal right; having a legal right does not guarantee a quick or simple remedy.

The details depend on the applicable law, the platform’s rules, the agreement and the facts. A disputed clone may raise questions about consent, ownership, parody, news reporting or other exceptions. Keeping evidence of the use and the relevant agreements can help clarify what happened.

The international picture, and what it signals for U.S. artists

Other countries are considering their own approaches. In Denmark, a proposed digital-replica law is expected to take effect in early 2027, according to the Danish culture minister, as reported by Complete Music Update and The Independent. Lithuania has also proposed a right intended to protect people from AI-generated voice clones, according to LRT reporting. These measures are proposals or pending developments, not settled law everywhere.

There are also efforts focused on music licensing, rather than a person’s voice or likeness as such. Sweden’s collecting society STIM has introduced an AI-rights approach based on writer-by-writer approval for licensing works. Independent publisher organizations IMPEL and IMPF have published principles asserting that AI-licensed songs should be valued at least equally to recordings. Those principles express an industry position; they are not a universal legal rule.

For U.S. artists, these developments show that the discussion is broader than whether a voice clone may be made. It also concerns who can approve a use, how permissions are recorded, and how creators are paid. The plain-language summary is that enforcement is running ahead of the law.

An empty vocal booth and microphone seen through glass, with subtle digital waveform reflections

A practical checklist for artists

Before signing an agreement or approving a synthetic use, consider whether it:

  • Limits AI training rights: Never grant open-ended AI training rights in a feature, sync or producer agreement.
  • Defines the terms: Spell out what “synthetic voice,” “digital replica” and “likeness” mean in the agreement.
  • Requires written approval: Require written approval before any AI-generated use of your voice or likeness.
  • Sets a time limit: Put a clear term limit on any license, rather than granting permission indefinitely.
  • Specifies payment: State who gets paid, how payments are calculated and when they are due.
  • Allows unused rights to be withdrawn: Make sure you can withdraw rights that have not been used, where the agreement permits.
  • Keeps proof of ownership: Retain original recordings, project files, dated drafts, split sheets, licenses and other records that can help prove authorship and ownership.
  • Registers your works: Register compositions and other eligible works with the relevant organizations and keep your registration details current.

These steps cannot replace legal protections or guarantee that a dispute will be resolved in your favor. They can, however, help make permissions clearer and preserve useful documentation.

Publishing

Songwriters and publishers should keep composition registrations and splits accurate, especially when discussing licenses for AI uses. Separately, “publishing” rights in a composition are not the same as royalties on a master recording. Artists who own or control their master recordings may also collect eligible SoundExchange royalties when their music is played on qualifying non-interactive digital services.

Radio-friendly talking points

  • 5-minute segment cue. What the stalled bill would change: Explain the federal voice-and-likeness right proposed by the NO FAKES Act, and clarify that the September 30 objection stopped unanimous-consent passage rather than producing a vote against the bill.
  • 6-minute segment cue. What artists can rely on today: Compare state laws such as Tennessee’s ELVIS Act with platform reporting policies and private contract terms. Emphasize that they are different kinds of protection.
  • 5-minute segment cue, The paperwork behind a voice: Discuss written consent, term limits, payment, withdrawal rights and preserving original recordings.
  • 7-minute segment cue. What other countries are considering: Summarize the pending Denmark and Lithuania proposals, then distinguish voice protection from music licensing approaches such as STIM’s approvals and the IMPEL/IMPF principles.
  • 5-minute segment cue, Voice rights versus music royalties: Explain the difference between a person’s voice, a musical composition and a master recording, including the separate eligibility rules for SoundExchange royalties.

Power JDM The Beat is following this issue as proposed laws, platform policies and licensing practices continue to develop.