Who Gets Paid When AI Learns the Music?
Poetically Speaking | November 3, 2026
A singer spends years finding a voice. A producer shapes a sound through hundreds of small decisions. A session musician can leave a whole personality inside a few bars. When that work becomes valuable to an AI company, who gets to say yes, and who gets paid?
That question is becoming a political question with very practical consequences for music. Congress is debating new rights. Public officials are taking opposing sides in court. Musicians are pressing their employers over how AI licensing money should be shared.
On this U.S. Election Day, it is worth looking past celebrity endorsements and campaign playlists. The rules surrounding music can affect an artist long after the last political ad disappears. Today’s election will not decide a pending copyright case or settle a recording contract. It will help determine who writes and scrutinizes the next round of policy.
1. Your voice and your recording raise different questions
Start with a distinction that gets lost when every story is filed under “AI.” A digital imitation of someone’s voice raises questions about identity and permission. Training a model on a recording raises questions about the use of copyrighted work. A licensing agreement raises another question: how is the money divided? One answer does not automatically resolve all three.
The proposed NO FAKES Act of 2026, S. 4591, focuses on voice and visual likeness. Its supporters want a federal right against unauthorized digital replicas, with a process for removing them from online platforms. This is proposed protection, not a new right artists should assume is already in force.
The politics are more complicated than a simple party split. On September 30, Democrat Chris Coons and Republican Marsha Blackburn argued for passage. Republican Ted Cruz blocked that unanimous-consent request, saying the bill needed stronger protection for parody and political speech. Blackburn and Coons maintained that the legislation already protected those uses. The Congressional Record captures both arguments.
Artists have a stake in both sides of that line: meaningful protection from impersonation and enough freedom to make satire, criticism and challenging art. The wording matters as much as the promise.
2. The AI training fight has reached the courtroom
A separate dispute concerns what AI developers may use to build their models. In the copyright litigation involving OpenAI and Microsoft, the Department of Justice has supported the defendants’ fair-use position. On October 30, Representatives Jamie Raskin, Jerrold Nadler and Hank Johnson announced a brief challenging DOJ’s argument, including its reliance on national-security concerns. Their position is that Congress did not create a national-security exception to copyright law.
OpenAI, the company behind ChatGPT, presents the opposing position. In an October 30 statement, it highlighted submissions arguing that AI training can be transformative, support new creative tools and serve broader public benefits. Those are arguments from a company defending its practices and its supporters, not findings that settle the case.
The dispute discussed in those statements concerns journalism and other written works. It is not a final ruling about every use of recorded music. Still, the underlying issue is familiar to musicians: when someone builds a commercial tool using existing creative work, where should permission, payment and lawful reuse meet?
That is where politics touches the studio. The government’s position in a lawsuit can support a particular vision of the market, while lawmakers can challenge that vision. Courts must decide the legal questions. An election result is no substitute for that process.
3. A licensing deal still leaves the question of who gets paid
Even when a company secures a license, the performers behind a recording may have separate concerns about consent and compensation.
The American Federation of Musicians launched its Human Made, Human Paid campaign on October 20, calling for musicians to share in value generated by AI deals using their performances. The union frames its demands around consent, credit and compensation. Its campaign explicitly says the goal is to protect human creators as technology develops.
This is an immediate workplace issue. In an October 29 update, AFM said Sound Recording Labor Agreement talks were scheduled to resume November 3–5. The union listed AI safeguards, healthcare funding and payment systems among its bargaining priorities. These are the union’s stated priorities and account of negotiations; this article does not report a completed agreement.
The distinction matters to anyone who has played on a recording without owning the master. A deal involving a catalog and a deal that fairly compensates everyone who helped create that catalog are separate achievements. Musicians want a voice in how the second one happens.
4. Independent artists need a seat at the negotiating table
For independent creators, the policy debate also concerns whether they can negotiate together.
Representative Deborah Ross introduced the Protect Working Musicians Act of 2026, H.R. 8994, on May 21. The proposal would create a limited antitrust safe harbor for qualifying music creator-owners to negotiate collectively over licensing with dominant music-distribution platforms and generative-AI companies. It includes eligibility requirements and conditions; it is not a guaranteed royalty rate or an automatic payment program.
Ross describes the bill as a way to strengthen small creators’ position when dealing with much larger companies. That is the political choice at its center: how should the law account for the difference in negotiating power between an independent rights holder and a major platform? The bill remains a proposal in the records reviewed for this piece.
Owning your work gives you something to negotiate with. The rules about negotiation help determine what that ownership can do for you.
5. What artists and listeners can ask now
You do not need to become a policy expert to ask specific questions. Three are a useful start:
Permission: Does a proposal or agreement distinguish using a recording, training a model and making a replica of a person’s voice? What choice does the creator retain?
Payment: Who is included when money changes hands? Consider the songwriter, recording owner, featured artist and musicians who performed on the track rather than treating “the music industry” as one beneficiary.
Accountability: Can the people affected understand what was used, challenge an unauthorized use and see how compensation is calculated?
Those questions apply to lawmakers’ proposals, platform announcements and industry deals. Fans can ask them, too. Admiring an artist’s talent and caring about the conditions under which that artist works belong in the same conversation.
The work behind the sound
Music is full of human decisions that a listener may never see: a rewritten line, a breath held a little longer, a bass note placed just behind the beat. The people making those decisions deserve to be heard when the business around their work changes.
That is the political story worth following beyond Election Day. Who can give permission? Who has room to negotiate? Who shares in the value?
The industry keeps asking artists to bring something original. Its next set of rules should leave them with a voice, a choice and a way to earn.
Which part of this debate matters most to you: protecting a voice, knowing how creative work is used, or making sure the people behind the music are paid?
Reporting checked November 3, 2026. Linked sources include official legislative materials and statements from the participants. Proposed legislation, advocacy positions and bargaining demands are identified as such. This article is commentary and general information, not legal advice.
Comments