Executive Overview
The intersection of generative artificial intelligence and copyright law has reached a critical juncture in Europe, with German collecting society GEMA emerging as one of the most aggressive and forward-thinking defenders of creators’ rights. Having recently secured a landmark initial legal victory against AI music platform Suno in Germany, GEMA is now charting a comprehensive strategy that spans litigation, high-stakes legislative lobbying, and the pursuit of fair market licensing.
At the heart of this strategy is a profound concern over the economic survival of songwriters, composers, and music publishers. In an exclusive interview for the Music Ally Focus podcast, GEMA General Counsel Dr. Kai Welp articulated the organization’s core philosophy: while unauthorized training data scraping remains a major legal infraction, the true existential threat to the modern music industry is the "substitution effect"—the commercial displacement of human-made music by automated, low-cost generative AI alternatives.
As streaming services, advertising agencies, gaming companies, and background music providers increasingly look to generative AI to cut costs, the potential erosion of traditional licensing revenue streams poses a catastrophic risk to creators. To combat this, GEMA is campaigning for a dual-pronged regulatory and commercial framework. This model demands fair, upfront compensation for copyright material utilized during the training phase, alongside a novel, mandatory system of "output remuneration" to compensate authors when AI-generated compositions directly replace human works in commercial and public spaces.
Furthermore, GEMA has cast a critical eye on existing legislative frameworks, dismissing the European Union’s flagship Artificial Intelligence Act as fundamentally insufficient to protect the economic interests of artists. This report explores GEMA’s ongoing legal battles, its policy objectives, the complexities of major label AI licensing agreements, and the broader implications for the global music ecosystem.
Detailed Chronology: GEMA’s Legal and Strategic Campaign Against Generative AI
To understand GEMA’s current positioning, it is essential to trace the escalation of tensions between collective management organizations (CMOs) and generative AI developers over the past several years. As text-to-audio and text-to-music models advanced from experimental novelties to commercial powerhouses capable of producing radio-ready tracks in seconds, rights holders realized that the legal paradigms governing traditional copyright were woefully unprepared for the digital age.
The Rise of Generative Audio and the Copyright Vacuum
Between 2023 and 2024, platforms like Suno, Udio, and various text-to-audio generators captured global attention by allowing users to create fully orchestrated songs complete with vocals and lyrics via simple text prompts. While consumers reveled in the democratization of music creation, the music industry quickly recognized the foundational flaw in these platforms’ development models: to train models to understand melody, harmony, rhythm, and song structure, developers had ingested massive datasets comprising millions of copyrighted sound recordings and musical compositions, typically without explicit authorization, attribution, or financial compensation.
While international music publishers and trade bodies initially filed lawsuits in the United States—targeting unauthorized training practices under traditional copyright infringement and fair use doctrines—European collecting societies began evaluating how European Union copyright directives applied to these technologies.
The German Front: GEMA Takes on Suno
GEMA, which represents over 90,000 members in Germany (including composers, lyricists, and music publishers), took a decisive stand by initiating legal action against Suno. The collecting society argued that Suno’s unauthorized ingestion of copyrighted works to train its AI models violated European copyright protections.
In a major breakthrough for the European creator community, GEMA secured an initial favorable ruling against Suno in Germany. This victory established a critical legal precedent: AI developers operating within European jurisdictions cannot simply claim immunity under tech-friendly safe harbors or data-mining exceptions without properly addressing the rights of copyright holders.
However, Dr. Kai Welp and his legal team recognize that winning a court case against a single platform is merely the opening salvo in a protracted war. The legal victory against Suno provided GEMA with the leverage needed to elevate its demands from reactive litigation to proactive, systemic policy reform.
Supporting Context & Metrics: The Anatomy of the "Substitution Effect"
To fully grasp why GEMA is pivoting so aggressively toward output-based remuneration, one must examine the macroeconomic realities facing contemporary songwriters and publishers in the streaming and synchronization markets.
The Fragility of Modern Creator Economics
For decades, the financial bedrock of a professional songwriter’s career has relied on a mix of mechanical royalties, performance rights royalties from public broadcast and performance (live venues, radio, television, background music in retail and hospitality), and synchronization fees.
However, the streaming economy—while expanding overall recorded music revenues—has famously resulted in a fractured micro-royalty distribution model for underlying composition owners, where only a tiny fraction of creators earn a sustainable living from streaming alone. Consequently, ancillary revenue streams, particularly public performance and background music licensing, have served as vital financial lifelines.
Quantifying the AI Threat: The Substitution Effect
Dr. Welp’s primary warning centers on the "substitution effect"—an economic phenomenon where a technological innovation replaces an existing product or service because it offers comparable utility at a significantly lower cost.
Consider the vast market for commercial background music (often referred to as production music or library music) used in corporate videos, elevator tracks, retail stores, restaurants, video games, and low-budget independent films. Historically, businesses and content creators licensed these tracks from libraries or independent composers, generating steady performance and mechanical royalty streams for GEMA’s members.
With the advent of generative AI music tools, a marketing agency or retail chain can generate endless hours of bespoke, royalty-free background music tailored precisely to their mood, tempo, and duration requirements in seconds, entirely free of ongoing licensing fees.
[Traditional Workflow]
Content Creator -> Licenses Human-Made Music -> Generates Royalties for Authors/Publishers
[AI-Driven Substitution Workflow]
Content Creator -> Prompts Generative AI Tool -> Zero Ongoing Licensing Fees -> Economic Loss for CreatorsAs Dr. Welp noted during his Music Ally Focus podcast interview:
"The most important thing is that we get an output remuneration for the authors and for the publishers. The reason for that is that we believe that the real threat to authors is the substitution effect of AI. If people use AI music, for example, as background music, they do not have to pay a licence fee — and it’s likely that companies will do that. So, the music authors will lose a lot of money and we need a compensation for that."
Without a mandatory mechanism to capture value at the point of consumption or output, the widespread adoption of AI-generated audio threatens to eviscerate the market for functional, commercial, and utility-based music, starving the human creative workforce of essential income.
Official Statements and Policy Analysis
GEMA’s strategic blueprint is not limited to fighting platforms in court; it targets the legislative halls of Brussels and Berlin, demanding structural reforms to ensure that artificial intelligence growth does not come at the expense of cultural diversity and human labor.
The Dual-Track Regulatory Model
Dr. Welp outlined GEMA’s comprehensive framework, which rests on two non-negotiable pillars:
- Fair Licensing on the Training Side: AI developers must secure explicit, commercial licenses from rights holders before scraping and ingesting copyrighted catalogs to train their models. Retroactive or uncompensated data harvesting must be outlawed or heavily penalized.
- Output Remuneration on the Consumption Side: Recognizing that tracking every single AI-generated prompt is practically impossible, GEMA is advocating for a statutory or collective remuneration right applied to the commercial use or distribution of AI-generated music. This mirrors traditional private copying levies or public performance tariffs, ensuring that when AI music displaces human music in commercial environments, a compensatory pool of money is collected and redistributed to human authors and publishers.
Critique of the European Union’s AI Act
While the European Union has positioned itself as a global leader in AI governance with the passage of the landmark EU AI Act, GEMA’s legal leadership remains distinctly unimpressed regarding its protection of copyright holders.
During his interview, Dr. Welp pulled no punches, explaining why he believes the EU AI Act is "really not sufficient" in its current iteration. While the Act introduces transparency requirements for general-purpose AI models—forcing developers to summarize the copyrighted data used in their training sets—it fails to establish robust enforcement mechanisms, mandatory collective bargaining frameworks, or guaranteed remuneration rights for artists whose livelihoods are threatened by downstream market substitution. For copyright collectives like GEMA, transparency without economic recourse is an empty gesture.
The Opacity of Major Label AI Deals
Compounding the regulatory challenge is the fractured nature of the modern music marketplace. Over the past year, major record labels (Universal, Sony, and Warner) have begun striking proprietary licensing and partnership deals with various AI audio startups and tech giants.
While these high-profile deals provide financial windfalls for major recorded music divisions, Dr. Welp highlighted a troubling lack of transparency and clarity for collecting societies. Because these agreements are negotiated behind closed doors—often focusing primarily on master rights and sound recordings—publishing rights, songwriting interests, and collective management administration are frequently left in a regulatory gray area. Collecting societies struggle to determine how these private deals impact statutory licensing tariffs, public performance collections, and the broader body of independent songwriters they represent.
Future Outlook: What Lies Ahead for AI and Copyright Law
As GEMA looks toward the remainder of the decade, its roadmap is defined by a blend of aggressive litigation, legislative lobbying, and technological adaptation. The battles being fought in German courts and European policy forums will establish a global precedent for how the creative industries navigate the artificial intelligence revolution.
1. The Push for Statutory Output Levies
Expect GEMA, alongside its international sister societies (such as ASCAP and BMI in the US, SACEM in France, and PRS for Music in the UK), to intensify lobbying efforts for statutory remuneration models. Just as radio broadcasters and public venues pay blanket licensing fees regardless of the specific origin of the music played, policymakers will face mounting pressure to institute levies on enterprise-grade AI generation tools used in commercial settings.
2. Expanding Litigation Beyond Germany
Following its initial legal success against Suno, GEMA is well-positioned to serve as the legal spearhead for European creators. As AI platforms continue to scale operations globally, cross-border jurisdictional enforcement will become paramount. Other European collecting societies are closely watching GEMA’s playbook, preparing to launch coordinated lawsuits if tech platforms fail to negotiate collective licensing agreements.
3. The Quest for Industry-Wide Transparency
The opaque nature of major label AI partnerships must be addressed to prevent the marginalization of independent publishers and self-published songwriters. GEMA will likely push for greater regulatory oversight to ensure that private tech deals do not undermine the collective bargaining power of traditional rights management organizations.
Conclusion
GEMA’s legal and strategic campaign against unchecked generative AI represents a defining moment for the creative economy. By recognizing that the true danger of artificial intelligence extends far beyond unauthorized data scraping into the systemic displacement of human labor via the substitution effect, GEMA is charting a course toward sustainable coexistence.
As Dr. Kai Welp and his team continue to press for fair training deals and mandatory output remuneration, the global music industry watches closely. The outcome of these efforts will determine whether the age of artificial intelligence results in the commercial devaluation of human creativity, or whether a balanced ecosystem can be preserved where human authors are fairly compensated for the cultural foundations upon which modern algorithms are built.
