Hamad Bin Khalifa University

For the first Global Dialogue to be considered a success, it must achieve three primary outcomes that move beyond high-level rhetoric toward institutional inclusion: 1. Institutionalizing Polycentricity: The Dialogue should result in a formal commitment to a "polycentric" governance model. This means moving away from a monocentric standard, where a specific ethical framework is the default, to a model where diverse moral, cultural, and religio-social traditions, including those from the Global South, are recognized as foundational normative contributors to global standards. 2. A Roadmap for Epistemic Equity: Success requires a concrete action plan to address the "digital invisibility" of underrepresented languages and civilizations. This includes a commitment to large-scale digitization of non-Western cultural and scholarly archives to ensure that global AI models are trained on representative "ground-truth" data, thereby preventing "Techno-Orientalism" and algorithmic stereotyping. 3. Reframing Ethics as an Enabler: The Dialogue should produce a framework that moves from "Techno-Commercial Determinism" to "Ethics by Design." This entails shifting the narrative from ethics as a restrictive regulatory burden to ethics as a catalyst for global trust and beneficial innovation. By ensuring AI is culturally aligned from the design stage, we can foster smoother international adoption and market acceptance. 4. Formal Integration of Regional Hubs: Establishing clear operational links between the UN Secretariat (ODET) and regional Centers of Excellence. This ensures that the UN Global Dialogue is not a top-down process, but an integrated ecosystem where various entities serve as independent reference points and gateways for diverse perspectives.

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

• Safe and Trustworthy AI (4a): We view "trust" as a function of cultural alignment. Trustworthy AI must be "Pro-Human" and grounded in "Ethics by Design." Systems that do not respect local values will inherently be viewed as untrustworthy and unsafe by the global majority. • Interoperability (4d): For a global dialogue to work, governance approaches must be compatible. We focus on interoperability to ensure that regional ethical frameworks (like a polycentric Global South approach) are not isolated but are strategically integrated into the international governance architecture. • Open-source/Open data (4g): This is critical for achieving Digital Equity. Our "Global Heritage AI Project" relies on open-data principles to integrate underrepresented knowledge systems into AI training sets. Open models are the most effective way to ensure that the Global South provides a "Voice, not an Echo" in the AI revolution, preventing data monopolies from enforcing a monocultural worldview.

Our Global Alliance for AI Ethics (GAAIE), hosted by Hamad Bin Khalifa University (HBKU) and Qatar Foundation identifies two critical emerging issues that require urgent integration into the Global Dialogue: 1. Techno-Orientalism and Algorithmic Stereotyping: While "human rights" and "bias" are mentioned, there is a specific need to address Techno-Orientalism-the phenomenon where historical and cultural stereotypes are uncritically encoded into LLMs as "objective" information. This is not just a technical bias; it is an epistemic injustice that distorts the representation of entire civilizations. The Dialogue must specifically address how we "decolonize" training data through the inclusion of human-curated, scholarly-annotated datasets from diverse traditions. 2. Distributed Responsibility and the "Wisdom Gap": As AI systems become background machine interactions (AI-to-AI), the traditional concept of responsibility (who to blame) is breaking down. We need to explore Distributed Responsibility-a property spanning developers, deployers, and users. Furthermore, we must bridge the "Wisdom Gap"-the disconnect between the accelerating pace of technology and the capacity of our cultural and moral institutions to respond. The Dialogue should explore how traditional moral wisdom (e.g., the concept of rukhṣah or beneficial leniency) can be utilized to open up alternative, beneficial innovation routes rather than simply resorting to reflexive prohibition. 3. Material and Resource Justice: AI ethics is often treated as a virtual problem, but it has a physical footprint. The extractive mining for AI hardware and the energy/water consumption of data centers primarily impact the Global South. A truly "polycentric" ethics must include the environmental and economic justice of the AI supply chain as a foundational ethical concern.

In the Middle East and the broader Global South, the primary governance gap is the prevalence of a monocentric ethical discourse. Current AI standards are largely defined by Western-centric intellectual and engineering frameworks, which often fail to account for the diverse moral traditions and cultural nuances of the Islamic world and other non-Western societies. * Significant Challenges: Techno-Orientalism and Digital Invisibility: A critical challenge is the systemic bias embedded in LLMs, where historical stereotypes are encoded as objective truths. This is compounded by the "Data Availability" gap; since Arabic and other regional languages are underrepresented in training sets, our civilizations are rendered digitally invisible or "hallucinated" by AI, leading to a loss of trust and cultural erosion. Techno-Commercial Determinism: In our rapidly developing sector, there is a risk that "technologically doable" and "financially profitable" become the sole metrics for progress, treatng ethics as a retrospective regulatory burden rather than a foundational design requirement. * Significant Opportunities: Establishing a Polycentric Reference Point: Qatar, as a global bridge, is uniquely positioned to fill this gap. We see an opportunity to move from being an "echo" of international norms to a normative co-shaper. By centering a polycentric approach, we can provide the UN with a credible reference point for inclusive AI governance. Epistemic Equity through Innovation: Through the Global Heritage AI Project, we are transforming a challenge into an opportunity by digitizing underrepresented scholarly archives. This provides the "ground-truth" data necessary for culturally authentic AI. Ethics as an Innovation Catalyst: By pioneering "Ethics by Design," we are showing that cultural alignment is not an inhibitor, but a driver for market acceptance and global trust, ensuring the AI revolution is humanly beneficial and universally relevant.

The AI Dialogue is uniquely positioned to transition global AI governance from a state of "Western-centric universalism" to a model of genuine polycentric consensus. Its most vital role is to serve as the definitive platform where global norms are not merely disseminated, but co-shaped by the global majority. Specifically, the Dialogue can advance cooperation through: 1. Legitimizing Polycentricity: By formally recognizing that diverse moral, cultural, and religious traditions are foundational contributors to governance, the Dialogue can bridge the current "legitimacy gap." This ensures that AI standards are inclusive of the worldviews of the 8 billion people they impact, moving beyond a monocentric standard. 2. Addressing the "Wisdom Gap": The Dialogue can facilitate a structured exchange between technical experts and cultural/moral authorities. This helps synchronize the accelerating pace of technological development with the capacity of societal and moral institutions to provide guidance, shifting the focus from reflexive prohibition to "Ethics by Design." 3. Standardizing Epistemic Justice: The Dialogue should establish international benchmarks for data representation. By championing the integration of Global South heritage and low-resource languages into the AI ecosystem, the Dialogue ensures that international cooperation is built on a foundation of digital equity rather than digital colonialism. 4. Facilitating Reciprocal Strategic Value: The Dialogue acts as the high-level mediator where the interests of the AI industry, national governments, and civil society are aligned. It provides the "International Reference Point" needed to transform ethics from a regulatory burden into a catalyst for global trust and market acceptance.

The AI Dialogue should act as the normative anchor that integrates and scales existing sector-specific efforts. Specifically, it should build upon: • The UNESCO Recommendation on the Ethics of AI: As the first global normative framework, the Dialogue should utilize UNESCO's readiness assessment tools while adding value by creating a Global South Multilateral Gateway—such as the GAAIE—to ensure these norms are culturally aligned at the drafting stage. • The Global Digital Compact (GDC) and ODET: The Dialogue must serve as the primary deliberative body for the implementation of the GDC, ensuring that the work of the UN Independent International Scientific Panel on AI is interpreted through a polycentric lens that respects diverse moral traditions. • Sectoral Frameworks (e.g., WHO/ITU AI for Health): The Dialogue should connect with specialized consortia to ensure that ethical standards in high-stakes sectors like healthcare are interoperable across different cultural and legal jurisdictions. The Added Value of the AI Dialogue: The unique value of the Dialogue lies in its ability to provide Institutional Anchoring. While academic centers and regional alliances (like the GAAIE) produce vital research and "ground-truth" data, they require the Dialogue's Universal Podium to translate those findings into global policy. The Dialogue brings the "Voice, not the Echo" to the UNGA by providing a formal mechanism to integrate independent auditing (via initiatives like the Global AI Ethics Observatory) and knowledge preservation (via the Global Heritage AI Project) into the official UN governance architecture. Ultimately, the Dialogue transforms fragmented regional initiatives into a cohesive, polycentric global

To be effective, the AI Dialogue must move beyond a consultative format toward a "Networked Governance" model that integrates stakeholders based on functional expertise. We recommend a three-tiered structure: 1. Strategic Policy Tier (Intergovernmental): Focused on high-level normative alignment and the adoption of universal standards. 2. Expert Working Groups (Multistakeholder): Specialized units comprising academic ethicists, AI engineers, and legal scholars. These groups should lead the technical design of frameworks, such as the "Ethics by Design" blueprint. 3. Regional Gateway Hubs: The Dialogue should formally recognize regional alliances (like the GAAIE) as "Regional Nodes." These hubs contribute by synthesizing regional moral traditions and "ground-truth" data into a format that can inform global policy. Stakeholder Contributions: • Academia: Should provide comparative ethical analysis and independent auditing methodologies. • Industry: Must contribute to the "Open Data" pool and provide transparency into the "Data Selection" processes to help identify bias. • Civil Society: Should act as the primary monitor for societal impact, particularly regarding "Techno-Orientalism" and marginalized communities. The format should prioritize "Recursive Deliberation," where technical findings from the UN Independent Scientific Panel are periodically reviewed by regional hubs to ensure they are culturally aligned before being finalized in Geneva or New York.

Currently, global discussions are characterized by a "monocentric" discourse that overlooks the Global South, the Islamic world, and indigenous communities. This is largely due to "Digital Invisibility," where populations with low digital footprints or non-digitized scholarly heritages are treated as data voids. Underrepresented Perspectives: 1. Non-Western Moral Traditions: Ethical frameworks from the Islamic, African, and Latin American traditions are often ignored in favor of Western-centric utilitarian or deontological models. 2. Low-Resource Language Communities: Perspectives rooted in languages other than English are marginalized by the current data-heavy nature of AI training. Strategies for Inclusion: • Epistemic Equity through Digitization: The UN should support initiatives like our "Global Heritage AI Project" to fund the digitization of underrepresented cultural and scholarly archives. This transforms these communities from passive data subjects into normative contributors. • Regional Gateways: Inclusion cannot be achieved through individual invitations alone. The Dialogue must utilize regional "Reference Points" (like HBKU in the Middle East) to facilitate a "Voice, not an Echo" for their respective civilizations. • Participation Support: Providing "Compute Credits" and travel support to scholars from the Global South to ensure they have a physical and digital seat at the table.

To prevent "dialogue fatigue" and ensure actionable outcomes, we propose moving beyond traditional panels toward more dynamic, output-oriented formats: 1. "Interdisciplinary Sync-Sprints": Short, high-intensity workshops where AI developers and ethicists from diverse moral traditions work together to solve a specific "Clinical Dilemma" (e.g., AI in healthcare accountability). The goal is to co-create a technical patch or a normative guideline in real-time. 2. "Polycentric Policy Labs": Regional breakout sessions held on the sidelines of the Dialogue, where a single global policy is "stress-tested" against different moral frameworks (e.g., how a privacy standard maps onto communal values in the Global South). This highlights areas of consensus and reasoned disagreement. 3. "Algorithmic Auditing Roundtables": Sessions where "Red-Teams" from regional hubs present findings from the Global AI Ethics Observatory to technology developers. This creates a direct feedback loop between independent auditors and the industry. 4. Asynchronous Multilingual Platforms: Utilizing AI-enabled translation to host an ongoing digital forum where scholars can submit inputs in their native languages (e.g., Arabic, Swahili, Quechua). This ensures that the Dialogue is not restricted to English-speaking elites but is accessible to the world's diverse intellectual leadership. These formats shift the focus from a "Western Echo" to a productive polycentric collaboration, making the Dialogue a true engine for humane technology.

Our Global Alliance for AI Ethics (GAAIE) at Hamad Bin Khalifa University highlights three concrete platforms and methodologies implemented at Hamad Bin Khalifa University (HBKU) that offer scalable solutions for global AI governance: 1. The Global AI Ethics Observatory (Auditing and Assessment): This platform serves as a "Regional Node" that complements international efforts by providing independent ethical auditing and scoring of AI systems. Its distinctive feature is Comparative Ethical Analysis, where AI-driven dilemmas are stress-tested against diverse moral traditions (e.g., Islamic ethics, Ubuntu, Confucianism). This ensures that AI models are not only technically robust but culturally aligned, moving beyond a "one-size-fits-all" Western-centric standard. 2. The Global Heritage AI Project (Data Representation): To solve the challenge of Digital Invisibility, this initiative coordinates the large-scale digitization of underrepresented cultural and scholarly archives. By curating high-quality "ground-truth" datasets for low-resource languages-specifically starting with Arabic-it provides a concrete technical solution to Techno-Orientalism. This practice secures Epistemic Equity, ensuring that AI systems are trained on data that reflects the world's actual cultural and linguistic diversity. 3. Interdisciplinary "Sync-Sprints" (Operational Practice): We utilize an agile methodology that bridges the gap between traditional moral scholarship and AI engineering. By implementing quarterly "Sync-Sprints," we integrate "Ethics by Design," ensuring that ethical guardrails are architected into the development lifecycle from the outset, rather than being treated as a retrospective regulatory check. 4. Polycentric Policy Labs: This approach facilitates high-level mediation between the technology industry, national governments, and diverse moral authorities. By seeking Reciprocal Strategic Value, we transform ethics from an innovation inhibitor into a catalyst for global trust and market acceptance, fostering an ecosystem where technological progress remains rooted in human dignity and the public interest.