GEMA – Gesellschaft für musikalische Aufführungs- und mechanische Vervielfältigungsrechte

A successful first Global Dialogue on AI Governance would produce practical, internationally relevant outcomes rather than only high-level declarations. From GEMA's perspective, success would mean establishing a shared baseline that confirms AI governance must remain human-centred and rights-based, with clear respect for existing legal frameworks, including copyright and related rights. It should also recognise cultural diversity and the sustainability of human creativity as public-interest objectives. A second important outcome would be agreement that meaningful transparency is a precondition for trust, accountability, and workable markets. Creators, users, regulators, and other affected stakeholders need usable information about how AI systems are developed, how they operate, and how outputs are generated. Without such transparency, rights cannot be exercised effectively, compliance cannot be assessed, and lawful licensing models cannot develop at scale. Third, the Dialogue should clarify that responsibility must be allocated across the AI value chain. Developers, deployers, and intermediaries should bear obligations proportionate to their technical control, economic benefit, and capacity to prevent or mitigate harm. This is especially important in cross-border AI ecosystems, where fragmented approaches create uncertainty and enforcement gaps. Finally, success would mean creating a pathway from principles to implementation: continued multistakeholder engagement, support for international coherence, and concrete follow-up on transparency, accountability, and fair participation in AI-driven value creation. If the first Dialogue helps translate shared principles into practical governance approaches that support innovation while protecting rights and human creativity, it will have been a meaningful success.

• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Interoperability of governance approaches
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

We selected these four priorities because they best reflect GEMA's perspective as a collective management organisation representing composers, lyricists and music publishers in a rapidly changing AI environment. Transparency, accountability, and human oversight is the most immediate priority. For creators and rightsholders, meaningful transparency is essential to understand how AI systems are developed, how protected works may be used, and how outputs are generated. Accountability and human oversight are necessary to ensure compliance, trust, and effective remedies where harm occurs. Protection and promotion of human rights is equally important because AI governance should remain human-centred and rights-based. In our view, this includes safeguarding the rights and interests of creators, protecting human agency, and ensuring that technological development does not undermine the legal and ethical foundations of creative work. Social, economic, ethical, cultural, linguistic and technical implications of AI is a key area because AI is already reshaping cultural and creative sectors. For GEMA, urgent questions include the economic sustainability of human creativity, fair participation in AI-driven value creation, and the protection of cultural diversity in increasingly automated content environments. Interoperability of governance approaches is a priority because fragmented national or regional approaches create uncertainty and enforcement gaps, especially in cross-border AI ecosystems. Greater international coherence is needed to make governance more effective, predictable, and workable in practice. Together, these four priorities support an AI governance approach that enables innovation while protecting rights, cultural diversity, and the long-term sustainability of human creativity.

Yes. From GEMA's perspective, several important cross-cutting issues deserve more explicit attention. First, the sustainability of human creativity and cultural ecosystems should be recognised as a distinct governance concern. AI governance is not only about technical safety or compliance; it is also about preserving the conditions under which human creators can continue to produce cultural value. If AI-driven markets weaken fair remuneration, visibility, and bargaining power for creators, this can have long-term effects on cultural diversity and creative sustainability. Second, consent, licensing, and fair participation in value creation should be treated more explicitly. In the creative sectors, one of the most urgent issues is how governance frameworks address the use of protected works in AI development and deployment. Where AI systems rely on creative content, governance should support lawful use, practical licensing solutions, and fair economic participation for those whose works underpin AI value creation. Third, cross-border enforcement and allocation of responsibility remains an emerging challenge. AI systems are often developed, trained, deployed, and commercialised across multiple jurisdictions. Even where legal principles exist, enforcement can be difficult if responsibilities are unclear or fragmented across the value chain. Finally, authenticity, attribution, and the visibility of human creation are becoming increasingly important. Generative AI can blur the distinction between human and synthetic content, which may confuse audiences, affect trust, and reduce recognition for human creators. Governance discussions should therefore also consider how to preserve transparency about origin and protect the social and cultural value of human authorship. These issues cut across the listed themes and are highly relevant for ensuring that AI governance remains human-centred, rights-based, and workable in practice.

The governance gaps in the thematic areas we selected are already having significant effects on the music and creative sector in Germany, across the European Union, and internationally. From GEMA's perspective, the most significant challenge is the gap between rapid AI deployment and the still limited level of meaningful transparency, accountability, and enforceable responsibility across the AI value chain. Where protected works are used without clear consent, visibility, or fair remuneration, legal uncertainty increases and trust in the market is undermined. This particularly affects creators and rightsholders, whose ability to understand, license, and enforce their rights depends on practical transparency and workable governance frameworks. A further challenge is fragmentation across jurisdictions. Different national or regional approaches to copyright, transparency, and accountability create enforcement gaps in a cross-border AI ecosystem. In practice, this can weaken the position of creators and smaller market participants, while concentrating power in a small number of global technology actors. The increasing volume of synthetic content also raises concerns about attribution, authenticity, cultural diversity, and the long-term economic sustainability of human creativity. At the same time, there are important opportunities. Effective and internationally coherent AI governance can enable innovation while preserving trust, legal certainty, and broad participation. For the creative sector, this means creating conditions under which AI can support new forms of creativity and economic activity without undermining the rights and interests of human creators. Greater transparency, clearer allocation of responsibility, and more interoperable governance approaches would help lawful licensing models, fair participation in value creation, and more sustainable development of AI in the cultural sector.

The AI Dialogue can play an important role by helping to build a shared international baseline for AI governance in a rapidly evolving and highly interconnected global environment. Its main value lies in creating a space where governments, international organisations, technical experts, civil society, and affected sectors can exchange perspectives and move toward common principles that are globally relevant while remaining flexible enough to accommodate different legal and institutional contexts. From GEMA's perspective, the Dialogue can be especially valuable in reducing fragmentation. AI systems are often developed, trained, deployed, and commercialised across borders, while governance remains largely national or regional. This creates uncertainty, uneven enforcement, and opportunities for regulatory arbitrage. The Dialogue can help promote greater interoperability of governance approaches by identifying minimum shared expectations on transparency, accountability, human oversight, and respect for existing legal frameworks. The Dialogue can also help translate high-level principles into practical cooperation. This includes sharing implementation experiences, identifying governance gaps, highlighting sector-specific impacts, and encouraging approaches that are workable not only for large technology actors but also for creators, smaller market participants, and stakeholders in different regions and stages of digital development. Finally, the AI Dialogue can strengthen inclusion and legitimacy in global AI governance. International cooperation will be more effective if it reflects the lived experience of those affected by AI systems, including cultural and creative sectors. By bringing these perspectives into a sustained multistakeholder process, the Dialogue can help ensure that international AI governance supports innovation while also protecting rights, cultural diversity, and fair participation in AI-driven value creation.

The AI Dialogue should build on existing international frameworks and initiatives rather than duplicate them. It can add value by providing a universal UN-based space where governments, industry, civil society, and affected sectors can work toward more coherent and inclusive AI governance. From GEMA's perspective, its main contribution would be to connect principles with implementation, support greater interoperability across governance approaches, and ensure that creators and cultural sectors are meaningfully included in global discussions. This is particularly important in cross-border AI contexts, where fragmented approaches create uncertainty and enforcement gaps. In this regard, GEMA's AI Charter offers a sector-specific reference point for human-centred and rights-based AI governance in the creative industries, with a focus on transparency, responsibility, fair participation in value creation, and the sustainability of cultural ecosystems: https://www.gema.de/documents/d/guest/gema_ai_charter_november_2024-pdf

Different stakeholders should contribute according to their roles and experience. Governments can help identify areas for international convergence, international organisations can support coordination, technical experts can clarify implementation challenges, and civil society, creators, and affected communities can provide evidence on real-world impacts. Private sector actors should contribute practical information on development, deployment, risk management, and compliance. From GEMA's perspective, the Dialogue should be structured as an ongoing multistakeholder process, not a one-off event. It should combine high-level discussions with focused thematic sessions and written input, allowing broad principles to be translated into practical approaches. The format should also be inclusive and accessible. Participation should go beyond governments and major technology companies to include creators, cultural and creative sectors, smaller market participants, and stakeholders from different regions. This is essential to ensure that AI governance is credible, workable, and grounded in real-world experience.

In global discussions on AI governance, the most underrepresented voices are often those most directly affected by AI systems but least able to shape governance debates. This includes creators, workers, smaller market participants, civil society organisations, and communities from the Global South. Cultural and creative sectors are also often underrepresented, even though AI is already reshaping how creative works are used, valued, and monetised. From GEMA's perspective, creators and rightsholders need stronger representation. Their experience is essential for understanding how AI affects human creativity, cultural diversity, fair remuneration, and the sustainability of creative ecosystems. Smaller language communities and culturally specific markets should also be more visible, as global AI development can otherwise reinforce existing inequalities and favour dominant languages, platforms, and business models. In addition, participation is often uneven because meaningful engagement requires time, expertise, funding, and access to international policy processes. This can exclude stakeholders who are directly affected but lack institutional capacity. These voices could be included through more accessible and structured participation formats: open consultations, written submissions, regional consultations, stakeholder roundtables, and support for travel or remote participation. The process should also include targeted outreach to underrepresented sectors and communities, including creators, collecting societies, grassroots civil society groups, and representatives from smaller and developing markets. A more balanced dialogue requires not only open access in principle, but active measures to ensure that participation is diverse, informed, and meaningful in practice.

A practical format would be a combination of short panel discussions and focused stakeholder exchanges. From GEMA's perspective, the Dialogue does not need overly complex formats. It would already be very useful to have well-structured panels with balanced participation from governments, international organisations, industry, civil society, and affected sectors such as the cultural and creative industries. GEMA could, for example, contribute by sending a representative to a panel in Geneva to bring the perspective of creators and collective management organisations into the discussion. In addition, brief thematic breakout sessions or moderated roundtables could help gather more specific input on key issues such as transparency, accountability, and cross-border governance. Written submissions before and after the meeting would also help ensure that the discussion is informed and leads to follow-up. Overall, the most effective format would be one that is simple, inclusive, and focused on practical exchange rather than overly ambitious structures

From GEMA's perspective, effective AI governance should build on practical solutions that work across both large-scale and highly granular uses of creative content. One important example is GEMA's AI Charter, which sets out a human-centred and rights-based approach for the creative sector, with a focus on transparency, responsibility, fair participation in value creation, and the sustainability of cultural ecosystems: https://www.gema.de/documents/d/guest/gema_ai_charter_november_2024-pdf A second example is evidence-based policymaking. GEMA's study "AI and Music", developed together with SACEM, is based on input from more than 15,000 creators and shows how strongly affected stakeholders value transparency, consent, and fair remuneration. This kind of evidence helps ensure that AI governance reflects real-world impacts, not only technical or commercial perspectives: https://www.gema.de/en/news/ai-study At the same time, practical governance also depends on effective enforcement and legal clarity. In this regard, experience from ongoing litigation against AI providers (https://www.gema.de/en/news/ai-and-music/ai-lawsuit) in a cross-border context highlights the need for clearer rules on transparency, responsibility, and the lawful use of protected works in AI systems. From our perspective, collective management organisations can also be part of the solution. Collecting societies such as GEMA are designed to balance interests between creators, users, and commercial actors through licensing. They have long experience in managing rights in ways that are both scalable and practical, including where uses are mass, cross-border, and difficult to manage individually. This is especially relevant in AI, where governance must address both very large systemic questions and highly fragmented, small-scale uses of protected works. Licensing frameworks and collective rights management can help bridge this gap by offering workable mechanisms for consent, remuneration, and legal certainty. For GEMA, the most promising approaches are therefore those that combine high-level principles with practical implementation tools, especially in the areas of transparency, accountability, licensing, and fair participation in value creation.