A union and the major labels wrote AI consent into a contract

SAG-AFTRA and the major labels reached a 2024 sound-recordings agreement requiring consent and minimum pay before an AI voice replica ships.

Historical event

April 12, 2024

First source published

April 12, 2024

Site publication

September 18, 2026

What happened

On 12 April 2024, the union SAG-AFTRA and a negotiating committee representing Warner Music Group, Sony Music Entertainment, Universal Music Group and Disney Music Group announced a tentative successor agreement to the SAG-AFTRA National Code of Fair Practice for Sound Recordings, covering the period from 1 January 2021 to 31 December 2026. The union's Executive Committee had approved the deal two days earlier, on 10 April, and sent it to members for ratification. The labels and the union published the announcement jointly, with each side's own statement attached.

What the documents say

The joint announcement, hosted on Universal Music Group's own newsroom, quotes SAG-AFTRA National Executive Director and Chief Negotiator Duncan Crabtree-Ireland describing the deal as establishing, for the first time, 'collective bargaining guardrails assuring singers and recording artists ethical and responsible treatment in the use of artificial intelligence'. The text is specific about mechanism, not just intent: the agreement confines the terms 'artist', 'singer' and 'royalty artist' to humans, and requires 'clear and conspicuous consent, along with minimum compensation requirements and specific details of intended use' before a label releases a sound recording using a digital replica of an artist's voice. A separate account of the same period, RIAA's record of an April 2024 Senate hearing, shows the same industry push continuing into public policy weeks later, with FKA twigs testifying that unconsented use of a voice 'leaves me very vulnerable'.

Why it matters for makers

A collective bargaining agreement is not copyright and it is not a statute; it binds signatory labels and the union's members to each other, not the wider world. Its mechanism is contractual gatekeeping at the point of release: a signatory label cannot release a record built on a digital vocal replica without consent, disclosure of intended use and minimum pay attached beforehand. For a producer working inside that system — a songwriter, session player or engineer on a major-label act — an AI vocal pass on a session now carries a paperwork trail before it can ship.

What to check before you use it

The joint announcement describes a tentative agreement awaiting member ratification; nothing in the document confirms a final ratified text, and this record does not establish whether ratification has since occurred. A maker relying on this protection should ask a client or label directly whether the final ratified terms match the April 2024 summary, since a tentative agreement can change before members vote. This is an editorial caution beyond what either cited document states.

• Is the recording covered by this agreement, or is the label a non-signatory?
• Has consent for a digital vocal replica been documented in writing, with intended use specified?
• Does the session paperwork show minimum compensation attached to the AI use, separate from the ordinary session fee?

The agreement's significance is structural rather than technical: it is the labour side of the same problem the ELVIS Act and the NO FAKES Act address through state and federal law, reached at a bargaining table instead of a legislature, and covering only the artists and labels who signed it.

Sources & reading trail

The joint press release quoting SAG-AFTRA and describing the consent, compensation and human-only definitions.

Source published: 12 April 2024 · Retrieved: 16 September 2026

Contemporaneous record of the wider April 2024 industry and artist push for consent-based AI voice protections.

Source published: 2 May 2024 · Retrieved: 16 September 2026

Papers, terms and official documents establish the record; the maker reading and the checks are Signal to Song editorial analysis. This retrospective draft does not imply the site published on the event date.

Continue reading

• A federal bill would give Americans a property right in their voice
• Tennessee made voice a protected property right against AI clones
• A voice clone’s consent has to satisfy a statute and a platform
• Browse the complete the archive

Published September 18, 2026, not on the date of the event described.