Neubauer Legal / AiNJA – AI Governance & Regulatory Strategy

A successful first Global Dialogue on AI Governance should move beyond high-level principles and deliver concrete, interoperable pathways for implementation. First, it should establish a shared understanding of interoperability as a core objective of AI governance – not merely technical compatibility, but alignment across legal, regulatory and institutional frameworks. Without this, fragmentation will undermine both innovation and protection. Second, the Dialogue should result in the identification of a limited number of actionable governance building blocks. These could include baseline transparency standards, accountability mechanisms, and human oversight requirements that can be adapted across jurisdictions while maintaining functional equivalence. Third, the Dialogue should initiate a structured process for translating the Scientific Panel's findings into policy-relevant outputs. This requires bridging the gap between technical expertise and regulatory design. Fourth, it should create continuity. Rather than a one-off exchange, the Dialogue should lead to ongoing working structures or thematic tracks that enable sustained cooperation between stakeholders. Finally, success should be measured by usability. Outputs should not remain abstract but support policymakers, organizations, and practitioners in operationalizing AI governance in real-world contexts. From my research on AI governance and global regulatory frameworks, interoperability emerges as the key condition for making governance both effective and scalable. The Dialogue offers a unique opportunity to move from fragmentation towards coordinated implementation.

• Interoperability of governance approaches
• Transparency, accountability, and human oversight
• Protection and promotion of human rights
• Safe, secure and trustworthy AI

The selected priorities reflect the structural conditions required for effective and globally coherent AI governance. Interoperability of governance approaches is central. Current regulatory developments risk creating fragmented and incompatible systems. Without interoperability, even well-designed frameworks cannot scale or interact effectively across jurisdictions. Transparency, accountability and human oversight form the operational core of governance. They translate abstract principles into enforceable and auditable mechanisms. Without these elements, governance remains declaratory rather than effective. The protection and promotion of human rights provide the normative foundation. AI systems increasingly affect fundamental rights, including autonomy, expression, and economic participation. Governance must ensure that these impacts are addressed consistently across legal systems. Safe, secure and trustworthy AI is essential to maintain public confidence and ensure that technological deployment does not create systemic risks. However, trustworthiness should not be treated as a standalone concept, but as the result of enforceable governance structures. Taken together, these priorities are not isolated themes but interdependent components of a functional governance architecture. From my research perspective, their alignment through interoperable frameworks is critical to avoid regulatory fragmentation and to enable meaningful international cooperation. This approach supports the transition from principle-based discussions to implementable governance models, which should be a central objective of the Global Dialogue.

Yes. A key cross-cutting issue not sufficiently captured by the listed themes is the question of how to operationalize interoperability in practice. While interoperability is recognized as a thematic area, its systemic implications go beyond a single category. It requires the development of translation mechanisms between different regulatory models, governance philosophies, and levels of enforcement. Without such mechanisms, even well-aligned principles may lead to fragmented implementation. Another emerging issue is the need to distinguish more clearly between principle-based governance and operational governance. Many existing frameworks articulate high-level values, but lack concrete pathways for implementation, monitoring, and enforcement. Bridging this gap is essential to ensure that governance is not only normative but functional. In addition, the interaction between AI governance and existing legal domains-such as intellectual property, data protection, and liability regimes-remains underdeveloped. These intersections are critical, as AI systems do not operate in isolation but within complex legal ecosystems. Greater attention should be given to how governance frameworks can be integrated rather than layered. Finally, there is a growing need to address the concept of human contribution and responsibility in AI systems more explicitly. As AI systems become more autonomous, questions of authorship, accountability, and attribution require clearer and internationally compatible approaches. From a research perspective, these cross-cutting issues point to the necessity of moving towards more coherent and interoperable governance architectures. Addressing them early could significantly enhance the effectiveness and sustainability of global AI governance efforts.

In a European context, significant progress has been made in establishing comprehensive AI governance frameworks, most notably through the EU AI Act. This represents a major step towards structured regulation and risk-based oversight. However, key governance gaps remain. One of the most pressing challenges is the emerging fragmentation between jurisdictions. While the EU is advancing detailed regulatory approaches, other regions are developing more flexible or sector-specific models. This divergence creates increasing complexity for organizations operating across borders and risks undermining the effectiveness of governance efforts. Another challenge lies in the transition from regulatory design to practical implementation. Organizations, particularly small and medium-sized enterprises, face difficulties in translating legal requirements into operational processes. This highlights a broader gap between principle-based regulation and actionable governance. At the same time, important opportunities are emerging. The European approach provides a valuable reference model that can inform global discussions. If aligned with other frameworks through interoperable mechanisms, it could contribute to the development of globally compatible governance structures. Furthermore, there is growing awareness across sectors of the need for accountability, transparency, and human oversight. This creates a favorable environment for advancing more structured and enforceable governance models. From my perspective, the central opportunity lies in bridging fragmentation through interoperability. Aligning different governance approaches—not by uniformity, but by functional compatibility—will be critical to ensuring that AI governance becomes both effective and scalable across jurisdictions.

The AI Dialogue can play a critical role as a coordination layer between existing governance efforts, rather than as an additional forum for principle-setting. Its primary contribution should be to enable interoperability between different national and regional governance approaches. This includes facilitating mutual understanding of regulatory models, identifying areas of functional equivalence, and supporting the development of mechanisms that allow different systems to interact without requiring uniformity. Furthermore, the Dialogue can act as a bridge between technical expertise and policy implementation. By translating insights from the Scientific Panel into structured policy options, it can help ensure that governance discussions lead to actionable outcomes rather than remaining at a conceptual level. Another key role lies in fostering continuity. International cooperation on AI governance requires sustained engagement beyond single events. The Dialogue could establish ongoing working tracks or thematic clusters that enable stakeholders to collaborate on specific governance challenges over time. Finally, the Dialogue should promote a shift from principle-based alignment towards operational alignment. This involves focusing on how governance can be implemented, monitored, and enforced across jurisdictions in practice. From my perspective, the Dialogue's added value lies in its ability to reduce fragmentation by enabling structured cooperation and interoperability. If designed accordingly, it could become a central platform for advancing coherent and globally compatible AI governance.

The AI Dialogue should build upon existing international and regional initiatives, while focusing on connecting them rather than duplicating efforts. Relevant frameworks include the OECD AI Principles, UNESCO's Recommendation on the Ethics of Artificial Intelligence, and regional regulatory approaches such as the EU AI Act. In addition, technical and standardization efforts, as well as emerging governance practices in different jurisdictions, should be taken into account. However, these initiatives currently operate largely in parallel. While they share common principles, their implementation pathways and enforcement mechanisms differ significantly. This creates a growing need for coordination and alignment. The added value of the AI Dialogue lies in its potential to act as an integrative layer across these efforts. Rather than developing new standalone principles, it should focus on identifying overlaps, gaps, and points of divergence between existing frameworks. A key contribution would be the development of interoperability-oriented tools or reference models. These could support policymakers and organizations in navigating different regulatory environments and enable a more coherent application of governance standards across jurisdictions. In addition, the Dialogue could highlight practical use cases and implementation experiences, helping to translate existing principles into operational governance models. From my perspective, the most important opportunity is to move from parallel governance initiatives towards a more connected and interoperable ecosystem. The AI Dialogue is uniquely positioned to facilitate this transition.

Different stakeholders contribute distinct but complementary perspectives to AI governance, and the Dialogue should be structured to make these contributions both visible and actionable. Member States can provide regulatory frameworks and policy priorities, while the private sector contributes practical implementation experience and insights into operational challenges. Academia and the technical community offer analytical depth and methodological rigor, and civil society brings essential perspectives on societal impact and fundamental rights. To ensure meaningful contributions, the Dialogue should move beyond sequential statements and adopt a more structured format. This could include thematic working groups or tracks aligned with key governance challenges, enabling stakeholders to engage in focused, solution-oriented discussions. In addition, mechanisms for synthesizing inputs are essential. Contributions should be systematically captured, compared, and translated into structured outputs, such as policy options or governance models. The Dialogue should also incorporate continuity by establishing ongoing collaboration formats, rather than relying solely on single events. This would allow stakeholders to build on each other's contributions over time. From my perspective, the effectiveness of stakeholder participation depends not only on inclusiveness but on structured interaction. Well-designed formats can transform diverse inputs into coherent and actionable governance outcomes.

Several perspectives remain underrepresented in global discussions on AI governance. First, there is a lack of representation from small and medium-sized enterprises (SMEs). While large technology companies are often well represented, SMEs face specific challenges in implementing governance requirements and are critical for understanding practical feasibility. Second, practitioners working at the intersection of law, technology, and organizational implementation are often missing. Their experience is essential for bridging the gap between regulatory design and real-world application. Third, voices from regions with emerging or developing AI ecosystems are still underrepresented. Ensuring broader geographic participation is crucial for avoiding governance models that are not globally applicable. In addition, creators and professionals in fields such as the cultural and creative industries are often not sufficiently included, despite being directly affected by AI systems. To address these gaps, participation should be actively facilitated. This includes lowering access barriers, providing structured opportunities for targeted stakeholder groups, and ensuring that contributions are meaningfully integrated into outcomes. From a governance perspective, inclusiveness must go beyond representation. It requires mechanisms that ensure that diverse perspectives influence the design and implementation of governance frameworks.

To foster meaningful and dynamic engagement, the AI Dialogue should move beyond traditional plenary formats and adopt more interactive and outcome-oriented approaches. One effective format could be structured thematic labs or working sessions, where stakeholders collaborate on specific governance challenges. These sessions should be designed to produce tangible outputs, such as draft frameworks, policy options, or implementation models. Another approach is the use of comparative case discussions. By examining how different jurisdictions address similar governance issues, participants can identify commonalities, divergences, and potential pathways for interoperability. In addition, iterative formats could be introduced, where discussions build on previous sessions. This would enable deeper engagement and continuity, rather than isolated contributions. Digital tools could also support more inclusive participation by allowing stakeholders to contribute asynchronously, particularly those who cannot attend live sessions. Finally, structured feedback loops are essential. Inputs from stakeholders should be reflected in outputs, and participants should be able to see how their contributions influence outcomes. From my perspective, the key to effective engagement lies in combining inclusiveness with structure. Innovative formats should not only increase participation, but also enhance the quality and usability of the Dialogue's outputs.

Effective AI governance requires approaches that combine normative principles with operational implementation mechanisms. One relevant example is the development of risk-based regulatory frameworks, such as the EU AI Act, which provide structured categorization and differentiated obligations. This approach supports proportional governance and offers a scalable model for other jurisdictions. In addition, transparency and accountability mechanisms are increasingly being operationalized through documentation requirements, audit processes, and human oversight structures. These elements are critical for translating governance principles into enforceable practices. A further important development lies in the emergence of interoperability-oriented approaches. Rather than aiming for uniform regulation, these approaches focus on achieving functional compatibility between different governance systems. This enables coordination across jurisdictions while respecting regulatory diversity. From my own research on AI governance and global copyright frameworks, it becomes evident that effective governance requires the integration of legal, technical, and organizational dimensions. In particular, the distinction between semantic and syntactic elements of AI-generated content, as well as the emphasis on human contribution and accountability, can support more precise governance models in areas such as authorship, responsibility, and rights allocation. More broadly, promising practices are those that move beyond abstract principles and provide implementable structures. This includes governance models that can be adapted across jurisdictions, while maintaining core functional equivalence. Such approaches contribute to reducing fragmentation and support the development of coherent and scalable AI governance systems.