Antonella Marra (Universal Music Italy): “Only works created by human beings can generate copyright”

Leaders League: You’ve been at Universal Music for over 21 years, witnessing the industry shift from physical formats to streaming, and now to the AI era. From an IP perspective, which transition has been the most disruptive, legally speaking, and what did it teach you about protecting artists’ rights in times of radical change?  
Antonella Marra: If I had to single out one transition, I would say that streaming has been the most visibly disruptive, but also the one we can now assess with greater clarity. It is a phenomenon that has been shaping the industry since the early 2010s, so with more than fifteen years of experience behind us, we have developed a more defined and critical understanding of its legal and economic implications.

Streaming fundamentally changed how rights are exploited and how revenues are generated, requiring a complete rethinking of existing legal and business models.

One important lesson from that period is that technological breakthroughs are often accompanied, at least initially, by phases of legal uncertainty and imbalance. The early days of digital music were marked by widespread piracy and systematic infringement of rights – the Napster era being the most emblematic example – which put the music industry under significant pressure and forced it to react in a highly defensive way.

As for the current generative AI transformation, my perception is that it may prove to be even more disruptive than streaming – although it is still too early to fully assess its impact. What we can say is that the music industry has learned important lessons from the past. Unlike during the early days of streaming, the music industry today is not only defending its rights through litigation when necessary, but is also proactively engaging with new technology. It is already exploring licensing models and working towards building a more sustainable AI ecosystem.

Unlike during the early days of streaming, the music industry is not only defending its rights through litigation, but is also proactively engaging with new technology

In the case of Universal Music Group, for example, AI agreements are already in place with companies such as Udio, Spotify and Klay Vision. These collaborations are built on a very clear framework: giving fans the opportunity to use artists’ recordings and voices to create content while avoiding direct competition between AI-generated music and human-created music and ensuring maximum transparency and traceability. In particular, one of the key principles underlying this early era of licensing at Universal Music is the so-called ‘walled garden’: initial fan-facing AI products will only allow playback on the service and no option to export nor upload to streaming services.

Ultimately, across all these transitions, one principle remains constant: innovation should not come at the expense of artists and rights holders. Our role is to ensure that, regardless of the format or technology, artists retain control over their work and are properly recognized and compensated for its use.

Universal Music has been at the center of some of the most significant AI-related IP battles in recent years – from the TikTok licensing standoff to lawsuits against AI music generators. As an in-house legal professional, how do you approach the tension between defending existing rights and finding commercial frameworks that allow innovation to coexist with protection?
Universal Music can truly be seen as a pioneer within the global music industry when it comes to engaging with new technologies, as well as developing and refining business models – in streaming with the “Artist centric” model, and now increasingly in the field of AI – with the aim of delivering greater value not only to the industry, but above all to the artists we represent.

You mentioned TikTok, which is a good example of how Universal Music consistently positions itself at the forefront when it comes to both defending artists’ rights and setting the standards for a fair and balanced ecosystem. As is well known, at the beginning of 2024 there was a standoff of the renewal of the license agreement between Universal Music with TikTok that lasted few months and had significant consequences – including the temporary removal of UMG music from the platform. However, that dispute was ultimately resolved and led to a mutually satisfactory new agreement between the two companies, which today continue to collaborate in a constructive and productive way.

Over the past two years, the music industry has taken a very firm legal stance against the unauthorized use of copyrighted works for AI training. Major record companies, including Universal Music Group, have initiated litigation against platforms such as Udio and Suno, on the basis that large-scale, unlicensed training constitutes copyright infringement.

Our role is to ensure that, regardless of the format or technology, artists retain control over their work and are properly recognized and compensated for its use

At the same time, however, Universal has consistently emphasized that litigation is only one side of the strategy. From the outset, the company has also worked to build a legitimate ecosystem for generative AI – one based on licensing, collaboration, and artist participation. The company has actively developed programs and partnerships with technology players willing to operate within a licensed, consent-based framework, with the goal of creating a model that works for both creators and innovators. This is evident in early collaborations of Universal Music Group with YouTube with his Music AI Incubator program well as in more recent agreements such as the partnerships with the AI music service Klay Vision and Stability AI, a professional AI tool for musician and artists.

All these efforts demonstrate that there is not necessarily a contradiction between defending existing rights and enabling innovation. On the contrary, these are complementary objectives that can and must coexist. The industry is not only taking action to enforce rights where necessary, but is also actively exploring new licensing frameworks and business models that ensure AI can develop in a way that respects creators and rights holders. 

AI tools can now replicate an artist’s voice with alarming accuracy. Beyond traditional copyright, we’re entering territory that touches on personality rights, image rights, and even moral rights. How do you see the legal toolkit evolving to address this, particularly within the European framework?
The evolution is already underway, but the starting point must remain unequivocal: No AI use that involves or replicates an artist’s voice can legitimately take place without his explicit authorization. This is a non-negotiable standard.

In the European context there is already a strong legal framework protecting personality, privacy, and identity rights – from GDPR to national rules on image and personality rights. In Italy, for example, these protections are rooted in our Constitution and reinforced by civil code and copyright law provisions that prohibit the unauthorized use of a person’s image.

I believe we will see a progressive evolution of the legal toolkit, combining existing rights with more specific regulatory interventions – as we are already starting to see with the AI Act.

The EU AI Act for example introduces a crucial shift toward transparency. It requires that AI-generated or manipulated content, including deepfakes, be clearly disclosed in many contexts. This does not prevent deepfakes per se, but it aims to reduce deception – addressing one of their most harmful effects.

However, legislation alone will not be sufficient. Contractual frameworks and industry standards will play a key role in defining how consent is obtained, how uses are authorized, and how value is shared

In particular, consent-based licensing models will be key, allowing artists to retain control over their voice and likeness while enabling legitimate, authorized uses under clear conditions, and fair remuneration.

As your work involves drafting and negotiating contracts in a landscape where the rules are still being written – especially around digital and new business agreements – how do you structure deals today to account for uses and technologies that don’t yet exist, or that regulators haven’t yet addressed?
The objective is not to predict the future perfectly, but to build agreements that are resilient and adaptable. This long-term perspective is not just theoretical – it has very concrete implications in day-to-day work. In my role, I regularly deal with the real-life impact of how contracts were drafted decades ago, and whether they are capable of accommodating technological change over time.

We can already see the consequences of this. Take streaming as a simple example: the ability for users to access individual tracks, reorder them, and include them in playlists is now fundamental to how music is consumed. However, in the era of the physical formats such as LP and CD, the sequencing of tracks on an album was often a core artistic and contractual element – carefully curated and, in some cases, subject to artist approval. The idea that songs could be freely unbundled and recombined by listeners might have been contractually restricted, or at least never contemplated.

Drafting contracts in this environment therefore requires a careful balance between precision and flexibility. On one hand, we need to clearly define rights, obligations and permitted uses. On the other, we must ensure that agreements are technologically neutral and sufficiently forward-looking to capture new forms of exploitation as they emerge.

In practice, this means using broader, technology-agnostic language where appropriate, combined with mechanisms that allow for adaptation over time. The goal is to protect value today without unintentionally limiting opportunities tomorrow.

IP law in the music industry ultimately exists to protect human creativity. After more than two decades working at the intersection of law, business, and art, how do you keep that human dimension at the center of your work – and what would you say to young lawyers who want to build a career in this space?
Copyright law has always been about protecting human creativity, but I would say this principle has never been more relevant than it is today. With the rapid rise of generative AI, one of the very first and most important steps taken by creative industries has been to reaffirm a fundamental concept: only works created by human beings can generate copyright. Consider for example the Human Artistry Campaign, a project that has been underway since the early stage of widespread generative adoption specifically aimed at promoting and protecting human creativity. This is not just a legal position – it is a cultural and economic statement about the value we place on human originality and artistic expression.

In my day-to-day work, keeping that human dimension at the center means constantly remembering that the contracts we negotiate are not abstract legal instruments. They have a direct and lasting impact on artists’ lives as they shape how rights are exploited and how royalties are paid over time. Working at an industry leading company like Universal Music, I feel a strong responsibility to help define frameworks that are both robust and fair: commercially sound for the business, but also balanced and sustainable for the artists we represent.

For young lawyers interested in this field, my advice would be twofold. First, develop a strong legal foundation – it is essential. But equally important is to understand the creative and commercial dynamics of the industry. Curiosity, adaptability, and the ability to think beyond purely legal frameworks are key.

This is an industry that constantly reinvents itself, and it offers a unique opportunity to work at the intersection of law, business and culture.