Major labels are getting “massive corporate payouts” for AI licensing but “paying musicians zero”, says AFM’s ‘Human Made Human Paid’ campaign

Executive Overview

The intersection of generative artificial intelligence and the music industry has officially breached a critical threshold. As technology companies aggressively train their models on vast catalogs of recorded music, the foundational labor force that built those recordings is fighting back. The American Federation of Musicians (AFM)—representing tens of thousands of professional performers, instrumentalists, and session players across the United States and Canada—has launched a high-stakes global advocacy initiative: the "Human Made Human Paid" campaign.

At its core, the campaign is designed to establish an uncompromising industry standard for the digital age: explicit musician consent, proper attribution, and fair financial compensation whenever human art is harvested for AI development. More immediately, the campaign serves as a strategic public relations and organizing weapon for the union as it navigates deeply contentious, stalled collective bargaining negotiations with the major record labels for a renewed Sound Recording Labor Agreement.

The stakes could not be higher. While major music conglomerates—including Universal Music Group and Warner Music Group—publicly sound the alarm about generative AI threatening to dilute the market with synthetic knock-offs and displace human artists, those same corporations are quietly executing lucrative licensing deals with tech platforms. They are monetizing their historic master catalogs through massive corporate payouts, yet the session musicians who played the original instruments driving these algorithms are, under current models, receiving zero financial return.

This investigative report breaks down the anatomy of the AFM’s new campaign, examines the legal and contractual war zone currently unfolding in US courts, analyzes the widening ideological chasm between artists and major rightsholders, and evaluates what this pivotal labor standoff means for the future of recorded music.


Detailed Chronology: The Road to Stalled Negotiations and Legal Warfare

To understand how the music industry arrived at this precarious juncture, one must examine the timeline of negotiations between the AFM and the Recording Industry Association of America (RIAA) majors. The friction is centered on the Sound Recording Labor Agreement, a master contract that sets minimum pay scales, health and pension contributions, and working conditions for session musicians performing on commercial sound recordings in the United States.

1. Rounds One Through Four: A Contractual Impasse

Throughout the year, the AFM bargaining committee sat across the table from major label representatives, seeking to forge a modernized contract that reflects the realities of a digitally disrupted market. However, after four intensive rounds of bargaining, negotiations ground to an absolute halt.

The sticking point? Contractual guardrails regarding artificial intelligence. While legacy compensation models covered physical formats, streaming, and digital downloads, they possess a glaring, multi-billion-dollar loophole when applied to machine learning and generative AI training datasets. The union’s demands for explicit protections against algorithmic displacement have been met with corporate resistance, forcing the AFM to leverage public mobilization to break the deadlock.

2. Legal Escalation: The Courtroom Battle Over "New Use" Clauses

The hostility between the union and the major labels is not confined to the negotiation table; it has already spilled over into the federal court system. The AFM has initiated legal action in US courts against Universal Music Group and Warner Music Group over how the existing Sound Recording Labor Agreement applies to AI licensing agreements.

At the heart of this legal dispute is the traditional "new use" clause embedded in union contracts. Historically, this clause was designed to protect musicians when their recordings were repurposed for mediums beyond their original intent—such as a track recorded for a vinyl album being licensed for a television commercial or a Hollywood film. In those instances, musicians were entitled to secondary compensation.

The AFM argues that licensing a master catalog to an AI developer to train a generative model is the ultimate manifestation of a "new use." Therefore, participating musicians should share in the licensing revenues. Conversely, the major labels have deployed legal teams armed with technicalities and narrow contractual definitions, arguing that training neural networks does not constitute a "new use" under the historic language of the agreement, thereby exempting them from paying secondary royalties.


Supporting Context & Metrics: The Three Great Debates of Music and AI

To contextualize the AFM’s crusade, it is helpful to look at the broader ecosystem of music and generative AI. Industry analysts generally categorize the ongoing crisis into three distinct ideological and legal battlegrounds:

  1. The Obligations of AI Tech Companies: This debate centers on whether tech platforms must secure licenses from copyright holders before scraping and training models on copyrighted audio files. Across the music community, stakeholders are largely united here: unauthorized data scraping is viewed as systemic copyright infringement, and platforms must negotiate commercial licenses.
  2. The Obligations of Labels and Publishers: This is the precise arena where the AFM’s campaign operates. It questions how major rightsholders—having successfully forced AI companies to the negotiating table—distribute the resulting windfalls. While labels and publishers champion their own rights to monetize catalogs, they have routinely failed to trickle those revenues down to the creators and session players whose performances gave those catalogs their intrinsic value.
  3. The Marketplace Influx of AI Music: The third debate tackles consumer protection, platform moderation, and the ethical dilemmas surrounding the proliferation of deepfakes, voice clones, and purely synthetic tracks flooding digital service providers (DSPs) like Spotify and Apple Music.

The Emerging Schist: Creators Versus Rightsholders

While the music community initially stood shoulder-to-shoulder against rogue tech startups, the announcement of monetization deals between major labels and AI companies has fractured this fragile coalition. A profound division has opened between artists, songwriters, and session musicians on one side, and major labels and music publishers on the other.

  • Featured Artists: For mainstream recording artists, control over their voice, likeness, and master recordings largely depends on the leverage they hold in their individual record deals. While global superstars can negotiate consent clauses, developing and mid-tier artists remain vulnerable to retroactive contract interpretations by risk-averse labels.
  • Songwriters: The landscape for songwriters varies wildly by jurisdiction and representation. In the US and UK, major publishers are aggressively cutting publisher-level AI deals. While collecting societies like the UK’s PRS for Music attempt to carve out protections and manage collective income streams, individual writers often find themselves at the mercy of opaque publishing contracts.
  • Session Musicians: Without doubt, session players face the most precarious existential threat. Their instrumental performances—often buried inside legacy recordings—are being reverse-engineered by AI models to synthesize infinite, bespoke instrumental tracks, effectively eliminating the need to hire human musicians for future studio sessions. Because session players typically work on a work-for-hire basis without ongoing royalties, they stand to make zero from AI licensing deals unless protected by robust copyright law reforms or powerful union contracts.

Official Statements and Core Demands

The "Human Made Human Paid" initiative is underpinned by a clear, non-negotiable manifesto. The AFM has codified its requirements into three foundational pillars that must govern any future use of human audio within artificial intelligence models:

The Three Non-Negotiable Pillars

  1. The Explicit Right of Refusal: Musicians must retain absolute autonomy to permit or explicitly refuse the utilization of their recorded performances, unique playing styles, and artistry for AI training datasets. Consent cannot be presumed or buried in fine print.
  2. Transparent Attribution and Credit: Whenever human artistry or audio is integrated into machine learning models, creators must be fully, visibly, and transparently credited for their foundational contributions.
  3. Equitable Financial Compensation: Musicians must receive a fair, sustainable share of the economic value generated whenever their work powers AI systems, commercial licensing deals, or algorithmic generation engines.

Strengthening Contractual Safeguards

Beyond these three pillars, the union is pushing for ironclad contractual guardrails to prevent record labels from utilizing AI tools to systematically undercut musician agency, replace human labor in recording studios, or exploit legacy sound recordings without permission.

In a rallying cry published on the official campaign portal, the AFM stated:

"This initiative unites musicians worldwide in defence of creators’ rights and establishes a vital standard for the digital age: if art is human-made, the creator must receive fair compensation, proper credit and explicit consent before their work can be used or automated."


Future Outlook: Industry Implications and the Road Ahead

As the "Human Made Human Paid" campaign gains momentum, collecting endorsements from musicians, labor advocates, and ethical tech organizations across the globe, the pressure on major record labels is intensifying.

The outcome of this standoff will set a monumental precedent for the creative economy at large. If the AFM successfully forces the major labels to concede at the bargaining table and secures favorable rulings in federal court, it will establish a legal blueprint for other creative unions—spanning actors, writers, and visual artists—facing similar technological displacement.

Conversely, if major labels succeed in locking session musicians out of the AI economy, it could accelerate a race to the bottom, where human studio labor is systematically replaced by synthetic equivalents trained on the unpaid sweat and talent of generations past.

For now, the message from the American Federation of Musicians is unambiguous: human creativity is not a free data pool for corporate tech experiments. As the digital age accelerates, the AFM’s campaign ensures that the human heartbeat behind the music is neither forgotten nor exploited.

You missed