University of Quebec in Montreal, Quebec , Canada and ORBICOM Network

A successful first Global Dialogue on AI Governance would not be defined by declarations alone, but by the establishment of operational capacities for collective mediation and proof across stakeholders. First, success would mean the creation of a shared transdisciplinary framework capable of bridging fragmented perspectives—technical, social, political, and ethical. Building on the Theory of Quantum Information and Communication (TQIC), this implies recognizing AI as an informational and communicational phenomenon embedded across multiple layers of reality, from data infrastructures to collective consciousness (noosphere). Second, success would require the adoption of an operational design methodology for governance. The Design Communautique 6.0 (DC 6.0) offers such a lifecycle—from exploration to evaluation—enabling structured, participatory, and iterative co-design of AI systems and policies. A dialogue that does not translate into actionable design processes risks remaining symbolic. Third, and most critically, success depends on establishing mechanisms of "Proof by Operational Ontological Mediation" (POOM). Current governance frameworks rely heavily on indicators and principles but lack the capacity to demonstrate, in a traceable and systemic way, that AI systems are aligned with human values. POOM introduces a multi-layered validation approach (ontological, epistemological, axiological, praxeological), ensuring that decisions are not only justified but operationally mediated across contexts. Finally, success would include the launch of a permanent multi-stakeholder platform—such as the ORBICOM initiative on a Global Constitution for AGI—to sustain dialogue through experimentation, prototyping, and continuous evaluation. In short, success is achieved when dialogue becomes design, and principles become provable, mediated action.

Social, economic, ethical, cultural, linguistic and technical implications of AI;Interoperability of governance approaches;Safe, secure and trustworthy AI;Open-source software, open data and open AI models;

These four priorities are deeply interconnected and reflect the need to move from fragmented governance approaches toward an integrated, systemic model. Safe, secure and trustworthy AI cannot be achieved solely through technical robustness. Trust emerges from the capacity of systems to be understood, debated, and collectively validated. This is why transparency, accountability, and human oversight are essential: they provide the conditions for meaningful mediation between stakeholders, rather than unilateral control by institutions or corporations. However, these dimensions remain insufficient without interoperability of governance approaches. Today's landscape is characterized by regulatory fragmentation, incompatible standards, and asymmetries between regions. A transdisciplinary and communication-centered framework-such as the one developed through TQIC and Design Communautique 6.0-can facilitate alignment by enabling shared representations, co-design processes, and iterative validation across contexts. Finally, the broad category of social, economic, ethical, cultural, linguistic and technical implications is not a residual domain but the core of AI governance. AI systems actively reshape meaning-making processes, social structures, and cultural dynamics. Ignoring these dimensions leads to polarization, epistemic fragmentation, and loss of collective agency. Together, these priorities point toward a shift: from governing AI as a tool to governing AI as a complex socio-technical and communicational system embedded in society.

A critical missing dimension in current AI governance discussions is the absence of operational mediation and proof mechanisms. Most existing frameworks focus on principles (ethics, safety, rights) or instruments (regulation, standards), but they lack the capacity to demonstrate-across contexts and over time-that AI systems are effectively aligned with these principles. This creates a gap between intention and implementation. The concept of Proof by Operational Ontological Mediation (POOM) addresses this gap by introducing a process-based approach to validation. Rather than relying solely on static indicators, POOM enables the tracing of decisions across multiple layers: what exists (ontology), what is known (epistemology), what is valued (axiology), and what is done (praxeology). This approach is particularly relevant for complex AI systems operating in dynamic and uncertain environments. Another emerging issue is the fragmentation of reality itself in the digital age. AI systems contribute to the proliferation of incompatible narratives, reinforcing what could be described as a "fractured noosphere." Governance must therefore include mechanisms for epistemic convergence and collective sense-making. Finally, there is a need to recognize AI as part of a broader evolutionary continuum of information and communication. This implies moving beyond purely technical governance toward a model that integrates human, social, and planetary dimensions. Without these additions, AI governance risks remaining reactive, fragmented, and ultimately ineffective in addressing the systemic challenges ahead.

The current governance gaps in AI are already producing tangible effects across countries, regions, and sectors, particularly in knowledge economies such as Canada and within international academic and policy networks like ORBICOM. The most significant challenge lies in the fragmentation of governance approaches. Regulatory frameworks, technical standards, and ethical guidelines evolve in parallel but remain insufficiently interoperable. This leads to inconsistencies in implementation, reduced trust, and difficulties in scaling responsible AI practices across jurisdictions. In sectors such as higher education, public policy, and digital innovation, this fragmentation translates into uncertainty, duplicated efforts, and uneven capacities to evaluate AI systems. A second major challenge concerns the absence of operational validation mechanisms. While principles of transparency, accountability, and safety are widely endorsed, there is no widely adopted method to demonstrate, in a systemic and traceable way, that AI systems effectively comply with them. This gap creates risks of "ethics washing" and weakens public confidence. At the societal level, AI systems are intensifying epistemic fragmentation—through misinformation, polarization, and the multiplication of incompatible narratives—affecting the cohesion of what can be understood as a shared noosphere. However, these challenges also open significant opportunities. They create the conditions for the emergence of new governance paradigms based on communication, co-design, and mediation. Initiatives such as ORBICOM's work toward a Global Constitution for AGI illustrate the potential to develop transdisciplinary platforms that integrate diverse stakeholders in iterative design processes. By combining frameworks such as the Theory of Quantum Information and Communication (TQIC), Design Communautique 6.0, and Proof by Operational Ontological Mediation (POOM), it becomes possible to move from fragmented governance to coordinated, verifiable, and participatory systems of AI governance. In this sense, current gaps are not only risks—they are catalysts for systemic innovation.

The AI Dialogue can play a decisive role by shifting international cooperation from principle-based alignment to operational co-design and mediated proof across stakeholders. At present, most international efforts converge at the level of ethical guidelines or regulatory coordination, but lack mechanisms to translate these into shared, actionable processes. The Dialogue can address this gap by fostering a transdisciplinary governance architecture grounded in communication and systemic integration. From the perspective of the Theory of Quantum Information and Communication (TQIC), AI must be understood as a multi-layered informational and communicational phenomenon, spanning technical infrastructures, social systems, and collective meaning-making processes. International cooperation therefore requires more than coordination—it requires shared frameworks for understanding and co-evolving these layers. The AI Dialogue could become a global mediation space structured around iterative design processes such as Design Communautique 6.0 (DC 6.0), enabling stakeholders to collaboratively explore, model, prototype, and evaluate governance solutions. This would transform dialogue into a continuous cycle of collective intelligence and institutional learning. Most importantly, the Dialogue should advance the integration of Proof by Operational Ontological Mediation (POOM) as a foundational capability. This would allow participating entities to demonstrate, in a transparent and traceable manner, how AI systems and policies align with shared values across ontological, epistemological, axiological, and praxeological dimensions. In this sense, the AI Dialogue can evolve from a forum of discussion into an infrastructure of global governance—capable of producing not only consensus, but verifiable, scalable, and adaptive solutions.

The AI Dialogue should build upon and interconnect a diverse ecosystem of existing initiatives, including UNESCO's Recommendation on the Ethics of AI, the Global Digital Compact, OECD AI Principles, GPAI, and emerging national and regional regulatory frameworks such as the EU AI Act. It should also engage with academic and transdisciplinary networks, including initiatives such as ORBICOM, which emphasize communication, knowledge-sharing, and global capacity-building. These initiatives provide essential normative foundations and policy coordination mechanisms. However, they remain largely fragmented and often operate at different levels—ethical, regulatory, technical—without sufficient integration into a coherent operational framework. The added value of the AI Dialogue lies in its potential to function as a connective and generative layer across these efforts. First, it can enable interoperability not only of standards, but of governance processes, by introducing shared design methodologies such as Design Communautique 6.0. This would allow diverse actors to collaboratively translate high-level principles into context-sensitive implementations. Second, it can introduce a new paradigm of validation through Proof by Operational Ontological Mediation (POOM). By focusing on traceability and mediation across multiple dimensions of governance, the Dialogue could strengthen the credibility and accountability of existing initiatives, moving beyond declarative commitments. Third, the Dialogue can act as a platform for experimentation and prototyping, linking policy frameworks with real-world applications through iterative cycles of testing and evaluation. Finally, by integrating perspectives from the Theory of Quantum Information and Communication (TQIC), it can contribute to a more holistic understanding of AI as a socio-technical and communicational system embedded in global society. In doing so, the AI Dialogue would not replace existing initiatives, but amplify their coherence, effectiveness, and long-term impact.

Different stakeholders can contribute most effectively if the AI Dialogue is structured not only as a forum for exchange, but as an operational co-design process. Governments can provide regulatory direction and legitimacy; industry actors can contribute technical expertise and implementation capacity; academia can offer theoretical integration and critical evaluation; civil society can ensure inclusivity, contextual relevance, and ethical grounding. However, their contributions must be mediated through shared processes rather than parallel interventions. We recommend structuring the Dialogue around iterative design cycles inspired by Design Communautique 6.0 (DC 6.0): exploration of issues, collective framing, solution modeling, prototyping, and evaluation. This would allow stakeholders to move from abstract positions to collaboratively designed governance solutions. To support this, the Dialogue should integrate structured deliberation methods (e.g., dialogic design processes) and digital collaboration platforms enabling continuous interaction beyond formal sessions. A key addition would be the introduction of Proof by Operational Ontological Mediation (POOM) as a transversal validation framework. Stakeholders would not only express positions, but demonstrate how their proposals align with shared principles across ontological, epistemological, axiological, and praxeological dimensions. In terms of format, a hybrid structure is recommended: plenary sessions for convergence, thematic working groups for depth, and ongoing collaborative labs for experimentation. This would transform the Dialogue into a living governance infrastructure rather than a one-time event.

Current global discussions on AI governance tend to overrepresent technical, governmental, and corporate actors, while underrepresenting communities directly affected by AI systems. Among the most underrepresented are populations from the Global South, Indigenous communities, non-dominant linguistic groups, and actors from the social sciences, humanities, and communication fields. These groups bring essential perspectives on cultural diversity, social impact, meaning-making, and lived experience, which are often overlooked in predominantly technical or regulatory frameworks. In addition, future-oriented perspectives—such as youth, interdisciplinary researchers, and practitioners working at the intersection of human, social, and ecological systems—remain insufficiently integrated. From a TQIC perspective, this imbalance reflects a broader issue: AI governance is often confined to the "bit level" (technical systems), while insufficient attention is given to higher levels of meaning and collective consciousness (noosphere). This creates risks of epistemic fragmentation and misalignment with societal values. To address this, inclusion must go beyond representation toward active participation in co-design processes. Mechanisms such as participatory design workshops, multilingual platforms, and capacity-building initiatives are essential. Frameworks like Design Communautique 6.0 can enable meaningful inclusion by structuring collaboration across diverse stakeholders, while POOM ensures that contributions are not only heard but integrated and validated within governance processes. Inclusion, therefore, should be understood as a condition for epistemic robustness and governance effectiveness, not only as a normative objective.

To foster meaningful and dynamic engagement, the AI Dialogue should adopt formats that move beyond static consultations toward interactive, iterative, and evidence-based processes. One promising approach is the use of structured dialogic design methods, enabling participants to collectively map problems, identify priorities, and co-construct solutions. These methods can be enhanced through digital platforms that support real-time collaboration, visualization, and traceability of discussions. Building on Design Communautique 6.0, the Dialogue could incorporate "governance labs" or "co-design studios," where stakeholders engage in guided cycles of exploration, modeling, prototyping, and evaluation. These labs would allow participants to test governance scenarios in controlled environments, bridging theory and practice. Another innovative format is the development of simulation-based engagement, such as serious games or scenario platforms, where participants can explore the systemic implications of different governance choices. This aligns with the idea of collective learning and anticipatory governance. To ensure rigor and accountability, these formats should be complemented by mechanisms of Proof by Operational Ontological Mediation (POOM), enabling participants to trace how decisions are justified and operationalized across multiple dimensions. Finally, hybrid and asynchronous formats are essential to ensure global participation, allowing stakeholders from different regions and time zones to contribute meaningfully. Together, these approaches can transform the AI Dialogue into a dynamic space of collective intelligence, experimentation, and continuous governance innovation.

A growing number of initiatives demonstrate promising directions for effective AI governance, yet most remain partial, fragmented, or limited to specific dimensions (ethical, regulatory, or technical). What is needed are integrative approaches capable of linking these dimensions into coherent, operational systems. In this regard, the recently accepted third collaborative project of ORBICOM-Towards a Global Constitution for AGI-offers a concrete and innovative contribution. This initiative aims to develop a transdisciplinary governance framework grounded in communication, collective intelligence, and participatory design. It combines the Theory of Quantum Information and Communication (TQIC), Design Communautique 6.0, and Proof by Operational Ontological Mediation (POOM) to enable stakeholders to co-design, prototype, and evaluate governance mechanisms in a structured and iterative manner. Beyond principles, it seeks to operationalize governance through living laboratories, observatories, and iterative policy design cycles, providing a scalable model for global coordination. Complementary initiatives also provide important building blocks. The UNESCO Recommendation on the Ethics of Artificial Intelligence establishes a widely recognized normative foundation. The OECD AI Principles contribute to policy alignment among member states. The Global Partnership on AI fosters international collaboration on applied AI governance challenges, while regulatory efforts such as the EU AI Act provide concrete legal frameworks for risk-based oversight. More experimental approaches are also emerging, including AI-assisted deliberation systems (e.g., "Habermas Machine") and participatory governance platforms that enhance collective decision-making. The added value of the ORBICOM initiative lies in its capacity to connect these efforts into an operational ecosystem-moving from fragmented governance to integrated, participatory, and verifiable systems. Together, these examples suggest that the future of AI governance lies not in isolated policies, but in interconnected platforms capable of continuous co-design, mediation, and validation at a global scale.