Ministry of Culture [solgulf]

The Dialogue will succeed if it produces a governance framework that works in practice across cultural contexts—not just on paper in Geneva. Success means establishing principle-based governance standards rather than process-based compliance checklists. Principles like human authority over outcomes, traceable accountability, and explainability tied to decision authority can be implemented universally while allowing cultural flexibility in execution. Success means recognizing that trust is built differently across different governance cultures. In institutional-trust contexts, process compliance validates decisions. In personal-accountability cultures—common across the GCC, Asia, Africa, and Latin America—governance requires named human ownership of every consequential AI decision. A global framework must accommodate both. Success means preventing fragmented adoption. If the framework assumes one accountability model, countries will sign it officially but bypass it informally. Systems will exist on paper while decisions happen elsewhere—a pattern I have observed repeatedly in government AI initiatives across Saudi Arabia. Most critically, success means answering one question clearly: when an AI system informs a decision and something goes wrong, who is accountable? If the answer is "the AI" or "the process," trust will not form—regardless of documentation. The Dialogue should produce a framework where: No AI system can make a decision without a responsible human authority Every AI-informed decision has a clear, auditable chain of accountability Countries can implement these principles using governance models aligned with their cultural realities Adoption is measured not by signatures, but by whether systems are genuinely trusted and sustained If we achieve this, AI governance becomes a foundation for global trust. If we build another compliance framework that ignores cultural accountability models, we build something governments will adopt in theory but abandon in practice.

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

These four priorities address the structural challenges that determine whether AI governance frameworks succeed or fail in practice. Cultural and ethical implications matter because AI governance cannot be culturally neutral. Accountability models differ: institutional-trust cultures validate decisions through process compliance, while personal-accountability cultures-common across the GCC, Asia, Africa, and Latin America-require named human ownership. Ignoring this leads to frameworks adopted on paper but bypassed in practice. Interoperability of governance approaches is essential for global adoption without fragmentation. A successful framework establishes universal principles-human authority over outcomes, traceable accountability, explainability tied to decision authority-while allowing countries to implement them using governance models aligned with their cultural realities. This prevents the pattern I observe repeatedly: systems deployed with full compliance but quietly sidelined because trust never forms. Transparency, accountability, and human oversight are foundational. No AI system should make a decision without a responsible human authority. When something goes wrong, the question is not "was the process followed?" but "who approved this, who can override it, and who is accountable?" If the answer is "the AI," the system is ungoverned regardless of documentation. Human rights protection ensures AI serves human dignity and agency rather than displacing them. This includes the right to understand how decisions affecting individuals are made, the right to human review and override, and protection against algorithmic harm especially in high-stakes contexts like justice, healthcare, and public services. Together, these priorities build governance that works across contexts, scales globally, and earns trust rather than simply demanding compliance. Without them, we risk creating another international framework that looks comprehensive but fails where implementation meets cultural reality.

Yes. One critical gap: the assumption that accountability models are culturally universal. Current AI governance frameworks are built on process-based accountability-if the procedure is documented and followed, the decision is defensible. This works in institutional-trust cultures where systems carry responsibility. It fails in personal-accountability cultures where governance requires naming the individual who approved, can override, and is accountable for each decision. This is not a minor implementation detail. It determines whether AI systems are genuinely adopted or quietly bypassed. I observe this pattern repeatedly across government AI initiatives in the GCC: organizations deploy systems with full compliance documentation, but people build parallel decision processes because trust never forms when accountability is invisible. The gap cuts across all listed themes: Safe and trustworthy AI depends on cultural models of trust Transparency and accountability mean different things process transparency vs. personal accountability Human oversight requires clarity on who has authority, not just that humans are "in the loop" Interoperability fails when frameworks assume one accountability model The emerging issue: as AI governance moves from principles to implementation, countries will face a choice adopt frameworks that don't align with cultural accountability structures, or bypass them informally. The result is fragmentation disguised as compliance. The solution: establish principle-based governance (human authority over outcomes, traceable accountability, explainability tied to decision authority) with cultural flexibility in implementation. Countries can then build governance systems aligned with how accountability actually functions in their context whether institutional-trust or personal-accountability based. Without addressing this, we risk creating another international framework that looks comprehensive but fails where global standards meet local realities.

The GCC faces a governance gap that Western frameworks do not address: the disconnect between process-based compliance and personal accountability cultures. Government AI initiatives across Saudi Arabia are adopting international governance standards—documentation, audits, compliance frameworks—but implementation fails quietly because these standards assume institutional accountability. In the GCC, trust is built on personal responsibility. When an AI system makes a recommendation, the questions that determine adoption are: who approved this? Who can override it? Who is accountable if it fails? If those questions cannot be answered with names, not processes, the system is not governed regardless of documentation. Organizations respond by building parallel decision processes. The AI system exists officially; decisions happen elsewhere informally. Transformation programs continue, reports show compliance, but trust never forms and adoption never scales. This creates visible symptoms: high AI investment with low operational integration, sophisticated systems that executives bypass, and governance frameworks that look comprehensive but don't prevent accountability gaps. The Opportunity: Vision 2030 requires AI governance that actually works not frameworks adopted for international alignment but ignored in practice. This creates demand for governance models that preserve human authority, ensure traceable accountability, and align with how decision-making actually functions in this region. The GCC can pioneer culturally-grounded AI governance: principle-based standards (human authority over outcomes, explainable accountability, override mechanisms) implemented through governance structures that match regional realities. This would demonstrate that effective AI governance is not culturally monolithic and provide a template for other personal-accountability cultures globally. The gap is real. The opportunity is to close it in ways that inform global standards rather than simply adopting them.

The AI Dialogue can advance cooperation by building a framework that works globally—not by imposing uniformity, but by establishing principles flexible enough to accommodate cultural diversity in governance. Three specific roles: 1. Bridge institutional-trust and personal-accountability cultures. Current AI governance standards assume process-based accountability works universally. The Dialogue can establish principle-based governance human authority over outcomes, traceable accountability, explainability tied to decision authority—that countries implement using culturally-aligned structures. This prevents fragmentation disguised as compliance. 2. Create implementation pathways, not just principles. Many international frameworks articulate ideals but provide no guidance on culturally-grounded implementation. The Dialogue can document how different governance cultures operationalize the same principles offering countries proven models rather than requiring them to translate Western frameworks into local contexts alone. 3. Establish mechanisms for ongoing learning and adaptation. AI governance will evolve faster than treaty cycles allow. The Dialogue can create structures for continuous knowledge exchange: what worked, what failed, which governance gaps emerged in practice. This prevents the pattern where frameworks become obsolete between adoption and implementation. The cooperation opportunity: Countries like Saudi Arabia, working through Vision 2030 AI initiatives, are building governance models for personal-accountability cultures in real-time. Countries in Asia, Africa, and Latin America face similar challenges. The Dialogue can connect these efforts not as isolated national experiments, but as a collaborative laboratory for culturally-grounded AI governance. Success means countries cooperate not by adopting identical frameworks, but by sharing how universal principles work across diverse contexts. This creates interoperability without cultural uniformity the foundation for genuine global cooperation.

OECD AI Principles provide a foundation but assume institutional-trust governance. The Dialogue should retain the principles (human-centered values, transparency, accountability) while expanding implementation guidance for personal-accountability cultures. UNESCO Recommendation on AI Ethics recognizes cultural diversity but lacks operational frameworks. The Dialogue can build on this by documenting how different cultures operationalize ethical AI moving from aspiration to implementation. EU AI Act demonstrates detailed regulation but reflects European governance assumptions. The Dialogue should study it as one implementation model, not the template—showing how principle-based governance can achieve similar outcomes through different cultural structures. Regional initiatives in the GCC: Saudi Arabia's SDAIA is building AI governance aligned with Vision 2030 and regional accountability models. The UAE's AI strategy emphasizes responsible AI with cultural considerations. These efforts represent governance models for personal-accountability cultures—directly relevant to large parts of Asia, Africa, and Latin America. The Dialogue should connect these regional efforts to global frameworks. The added value the Dialogue brings: Cross-cultural governance translation. Existing initiatives operate within their cultural contexts. The Dialogue can bridge them showing how OECD principles, EU regulations, and GCC governance models all protect human authority and accountability through different mechanisms. Validation of culturally-grounded approaches. Regional initiatives often feel isolated from global standards. The Dialogue can validate that culturally-aligned implementation is not deviation from global norms it is how global principles work in practice. Prevention of governance fragmentation. Without coordination, we risk competing frameworks: Western process-based, GCC personal-accountability, Asian hybrid models. The Dialogue can harmonize these not by uniformity, but by showing they implement the same core principles through culturally appropriate structures. The opportunity is building on what exists not replacing it, but connecting it into a coherent global approach.

The Dialogue should structure participation to surface implementation experience, not just policy positions. Recommended structure: Multi-track engagement: Policy track for government representatives establishing principles Implementation track for practitioners sharing what works in practice across different cultural contexts Technical track for developers and researchers addressing operational challenges Civil society track ensuring affected communities inform governance design Problem-based working groups: Rather than theoretical discussions, organize around real governance challenges: "How do we ensure traceable accountability in AI systems?" "How do cultures with different trust models implement the same principles?" Working groups should include diverse stakeholder types addressing the same problem. Regional consultation rounds: Before Geneva and New York convenings, hold regional consultations in GCC, Southeast Asia, Africa, Latin America, and Eastern Europe. This surfaces governance models beyond Western frameworks and validates that regional approaches contribute to rather than deviate from global standards. Implementation case studies: Require participants to submit not just position papers, but documented cases: what governance approach was attempted, what worked, what failed, what was learned. This creates a knowledge base of how principles translate to practice across contexts. Ongoing contribution mechanisms: AI governance evolves faster than diplomatic cycles. Create platforms for continuous contribution between formal convenings—allowing stakeholders to share emerging challenges and solutions as implementation progresses. The format principle: Structure contribution to value lived experience as much as formal expertise. The practitioner who implemented AI governance in a Saudi ministry and observed quiet system bypass has knowledge as valuable as the policy expert who designed the framework. Format should capture both. Effective structure means diverse stakeholders contribute different knowledge types all necessary for governance that works in practice, not just on paper.

Underrepresented: Practitioners from personal-accountability cultures. Global AI governance discussions are dominated by voices from institutional-trust contexts primarily Western Europe and North America. This creates a dangerous assumption: that process-based accountability is universal. Missing are practitioners from the GCC, much of Asia, Africa, and Latin America who work daily with governance structures where accountability is personal, not institutional. These voices don't just represent different regions—they represent fundamentally different governance models that affect how AI systems are trusted and adopted. Specific gaps: Government AI implementation teams in personal-accountability cultures who observe patterns like system bypass, parallel decision processes, and compliance without trust knowledge critical for understanding why Western frameworks fail in practice elsewhere. Regional AI ethics boards and oversight bodies designing culturally-grounded governance that achieves the same protective outcomes through different accountability structures. Arabic, Mandarin, Hindi, Swahili-speaking AI communities whose governance discussions happen in languages other than English excluding their frameworks and terminology from global discourse. How to include them: Regional expert panels convened in Riyadh, Singapore, Nairobi, São Paulo—not just satellite events, but formal Dialogue components with equal weight to Geneva/New York proceedings. Documentation in multiple languages allowing submissions and participation in Arabic, Mandarin, French, Spanish, Hindi with professional translation ensuring ideas aren't filtered through English-language framing. Implementation-focused selection criteria that value practitioners over theorists, giving voice to people solving real governance challenges in diverse contexts. The inclusion principle: A global framework built primarily on institutional-trust governance models will be adopted globally but implemented fragmentarily. Including personal-accountability voices isn't diversity for its own sake—it's essential for building governance that works across the cultures that will need to implement it.

Governance labs over panel discussions. Traditional formats position papers, keynote speeches, roundtables—generate talk. Labs generate solutions. Proposed format: Cross-cultural implementation labs: Bring together practitioners from institutional-trust and personal-accountability cultures to solve the same governance problem in parallel. Example lab: "Design an AI approval workflow that ensures human authority and traceable accountability." Western teams design process-based solutions. GCC teams design person-based solutions. Both achieve the same protective outcome through different mechanisms. The lab documents both, showing cultural implementation flexibility within universal principles. Live governance case studies: Rather than discussing AI governance abstractly, examine real systems deployed in different countries. Saudi government AI system. EU public service AI. Singaporean healthcare AI. African agricultural AI. Multi-stakeholder teams analyze: How is accountability structured? Where does trust form or break? What governance worked and what was bypassed? Simulation exercises: Present participants with governance scenarios: "An AI system recommends budget reallocation. Regulators question the decision six months later. Using your governance framework, demonstrate who is accountable and how you would explain the decision." Teams from different cultures solve the same scenario revealing where frameworks provide clarity and where critical gaps remain. Bilateral implementation exchanges: Pair countries with different governance cultures to exchange practitioners for short-term implementation observation. Saudi practitioner embeds with EU AI governance team. EU practitioner embeds with SDAIA. They document: What assumptions does each model make? Where do cultural differences matter most? How can both achieve interoperability? Digital collaboration platforms: Between formal convenings, maintain platforms where practitioners share implementation challenges in real-time, receive peer input, and build a living knowledge base of what works across contexts. The innovation principle: Engagement should generate implementable knowledge, not just policy consensus.ĺ

Three approaches addressing the accountability challenge: 1. SDAIA's National AI Strategy (Saudi Arabia) Establishes AI governance aligned with personal-accountability culture and Vision 2030. Key practice: Every AI system deployed in government must identify a named executive owner responsible for decisions the system informs. This isn't bureaucratic overhead it's culturally-grounded trust architecture. The system cannot go live until three questions have documented answers: who approved deployment, who can override recommendations, who is accountable for outcomes. This prevents the pattern where AI systems exist officially but are bypassed informally because accountability is invisible. 2. Singapore's Model AI Governance Framework Provides principle-based guidance with implementation flexibility. Strength: recognizes that governance must adapt to organizational context while maintaining core protections. Demonstrates how universal principles (transparency, explainability, human oversight) can be implemented through different structures. Weakness: still assumes institutional accountability could be strengthened by explicitly addressing personal-accountability implementation. 3. EU AI Act's High-Risk Classification Creates differentiated governance based on decision impact. Critical insight: not all AI needs the same governance intensity. High-stakes decisions (justice, healthcare, critical infrastructure) require stricter human oversight and accountability chains. This prevents both over-regulation of low-risk systems and under-regulation of consequential ones. The pattern that works: Effective governance combines universal principles with cultural implementation flexibility. It answers "who is accountable?" before deployment, not after failure. It recognizes that trust is built differently across contexts and allows governance structures to align with how accountability actually functions. The gap remaining: No existing framework fully addresses interoperability between institutional-trust and personal-accountability governance models. The Dialogue can close this not by choosing one model, but by showing how both achieve the same protective outcomes through culturally-appropriate structures.