A Munich courtroom is set to deliver the first major court ruling on whether AI music generators like Suno illegally trained on copyrighted songs, a verdict that could reshape how the music industry and regulators worldwide approach generative AI. On July 31, Germany's GEMA (the country's equivalent of ASCAP and BMI combined) will present its case against Suno, marking a watershed moment: unlike American lawsuits that have dragged on for years fighting over what went into training datasets, both sides in Munich already agree the songs were used. The question now is whether that training constitutes infringement.

What Makes the Munich Case Different From U.S. Lawsuits?

The German case skips past the access and training disputes that have consumed American litigation. In the U.S., Universal Music Group and Warner Music Group settled with Suno, while Sony continues fighting, with both sides spending years arguing about what music was scraped and how. In Munich, that battle is already over. GEMA sued Suno in January 2025 over six specific songs: "Forever Young," "Mambo No. 5," "Daddy Cool," "Rasputin," "Big in Japan," and a sixth song well-known in Germany. Nobody disputes that Suno obtained these songs from YouTube and used them in training.

Instead, the Munich court is focusing on something American courts have largely avoided: whether the outputs themselves infringe on the original works. GEMA created test outputs by prompting Suno with the original song titles, lyrics, and styles, then presented those generated tracks as evidence of infringement. The court must now determine whether those AI-generated outputs constitute unauthorized reproduction and public distribution of protected works.

This distinction matters enormously. Germany has no "fair use" doctrine like the United States does. Instead, it has a text-and-data-mining exception, and the same Munich court chamber already ruled against OpenAI on this issue last November, holding that text-and-data-mining is meant for analysis, not for models that can regurgitate whole works. Critically, that ruling placed responsibility on the provider, not the user who prompted the system.

How Are Courts Evaluating AI Music Similarity?

The technical question at the heart of the case is deceptively simple: when an AI generates a song that sounds like an existing work, what exactly has been copied? Musicologists typically evaluate infringement using three criteria: access (did the creator hear the original?), similarity (do the works sound alike?), and protectability (is what's similar actually protected by copyright?).

In GEMA's test outputs, the court will examine whether the melodies, harmonies, and arrangements that emerged from Suno's generator match the original songs. Here's where the case gets interesting: GEMA's prompts apparently included no melody, rhythm, harmony, or arrangement instructions. Yet according to hearing reports, the melodies came back anyway. If those melodies are substantially similar to the originals, it suggests the model stored and reproduced protected musical expression. If they're generic or coincidental, GEMA's case weakens significantly.

Suno's defense rests on several arguments: the German court lacks jurisdiction, U.S. fair use protections apply, the outputs aren't recognizable as the original songs, the model stores only mathematical patterns rather than actual works, and users, not Suno, own the outputs. The last argument has already lost once in this same courtroom, making it a risky strategy.

What Could This Verdict Mean for the Music Industry?

The stakes extend far beyond Suno. If GEMA wins, it establishes that training AI models on copyrighted music without a license violates EU law, a significant precedent for European regulators and companies. However, it would say little about how U.S. courts might rule, since American copyright law differs substantially. If Suno wins, the text-and-data-mining question gets answered differently, potentially providing guidance for U.S. cases, though the core musicological question of output similarity would likely remain unresolved.

What the Munich verdict will not do is settle the question that American courts will eventually face: when an AI output resembles a real song, is the similarity rooted in protectable expression or in common musical vocabulary that copyright leaves free for everyone to use? That question is unavoidable and will likely determine liability for both AI music platforms and their users.

Steps to Understanding the Legal Landscape for AI Music

• Track the Munich Ruling: Follow the July 31 verdict closely, as it will be the first judicial treatment of AI musical memorization and regurgitation anywhere in the world, setting a template for how courts evaluate output similarity.
• Monitor Settlement Patterns: Watch whether the Munich outcome influences ongoing U.S. lawsuits from Sony and other rightsholders, or whether settlements emerge that might inoculate AI platforms from liability.
• Understand the Fair Use Question: Recognize that U.S. cases hinge on fair use doctrine, while EU cases rely on text-and-data-mining exceptions, meaning different jurisdictions may reach opposite conclusions on the same facts.

Why Suno's Business Momentum Matters Amid Legal Uncertainty

Despite the legal cloud, Suno has grown into a dominant force in AI music generation. The platform has attracted over 100 million users since its December 2023 launch, commanding roughly 69% market share in the music-generator space based on web traffic. In May 2026, Suno raised $400 million in Series D funding at a $5.4 billion valuation, more than doubling its $2.45 billion valuation from just six months earlier. The company now generates $300 million in annual recurring revenue from approximately 2 million paid subscribers, a conversion rate of about 2% of its free user base.

Suno's growth was turbocharged by a viral "text-to-song" trend on TikTok earlier in 2026, where users fed humorous messaging threads into the platform to generate songs. U.S. downloads of Suno's mobile app quadrupled during the trend's peak, and the company quickly shipped a feature to automatically convert text screenshots into lyrics. This hinted at a major use case: AI music as a tool for memes and social communication.

However, the company faces mounting challenges beyond litigation. In November 2025, Suno suffered a data breach that exposed information on more than 55 million users, including email addresses, phone numbers, physical addresses, and partial credit card data. While Suno confirmed a "security incident" in November, it did not dispute the scale of the breach reported by Have I Been Pwned, a security tracking database.

To build legitimacy with the music industry, Suno has hired veteran music executives, including Jeremy Sirota, former CEO of indie licensing agency Merlin, as chief commercial officer, and Paul Sinclair, a former Atlantic Records executive, as chief music officer. The company also signed a licensing deal with Warner Music Group in November 2025, allowing artists who opt in to have their names, images, likenesses, voices, and compositions used in AI-generated music in exchange for compensation.

The Munich verdict on July 31 will be the first judicial test of whether these licensing deals and fair use arguments can withstand scrutiny in court. Whatever the outcome, the ruling will likely accelerate settlement negotiations in the U.S. and force both AI platforms and rightsholders to confront the fundamental question of what copyright protects in an age of generative AI.