Independent Contributor

In my view, the first Global Dialogue on AI Governance would truly succeed if it delivered three practical outcomes instead of the usual diplomatic language. First, I'd love to see a short, straightforward "AI Commons Charter" — just a few clear, workable principles around verifiable safety testing, transparency for frontier models, and countries agreeing not to block each other's safe AI development. Nothing heavy-handed, just enough to keep innovation flowing across borders. Second, it should kick off real workstreams with actual deadlines: a shared public benchmark for serious risks like biological or cyber threats, developed together with researchers and companies, and a genuine capacity-building fund for the Global South that delivers compute, data, and training resources — not just promises. Third, I hope they openly reject treating every AI model as a danger. AI is an incredible general-purpose technology that could transform medicine, climate solutions, and energy. The Dialogue should champion responsible progress, not fear-driven restrictions that slow us down. If these three things happen — a simple charter, concrete benchmarks and funding, and a pro-innovation mindset — I'd consider the Dialogue a real success. Anything less would feel like just another talk shop. What do you think?

• Safe, secure and trustworthy AI
• Interoperability of governance approaches
• Transparency, accountability, and human oversight
• Open-source software, open data and open AI models

From my perspective, the four thematic areas that most urgently need active engagement and priority action in the Global Dialogue on AI Governance are: Safe, secure, and trustworthy AI; Interoperability of governance approaches; Transparency, accountability, and human oversight; and Open-source software, open data and open AI models. These four stand out to me because they directly tackle the biggest near-term risks while enabling faster, responsible progress. Safe and trustworthy AI should focus on evidence-based work on real catastrophic risks through shared benchmarks and verifiable testing, rather than broad rules that slow innovation. Interoperability of governance approaches feels especially urgent right now. We need to avoid a fragmented global patchwork of rules by agreeing on mutual recognition of safety certifications so safe AI can flow freely across borders. Transparency, accountability, and human oversight are key to building real trust through genuine openness about model capabilities instead of just promises. Open-source software, open data, and open AI models matter deeply to me. They accelerate scientific discovery, enable broader safety scrutiny, democratize access (especially for the Global South), and help prevent dangerous concentration of power. The other areas are important long-term, but these four feel most pressing for concrete outcomes rather than another talk shop. This approach supports pragmatic, pro-innovation governance that maximizes AI's benefits for humanity while managing real harms.

Yes, in my opinion, there is one major cross-cutting issue that the listed themes do not fully capture: the risk of regulatory fragmentation and the urgent need for practical mechanisms to keep AI development global and competitive. The listed themes focus heavily on safety, openness, human rights, ethics, and capacity - all important. However, they risk missing how quickly incompatible national rules (export controls on models and compute, divergent safety standards, or conflicting liability regimes) could splinter the AI ecosystem. This would raise costs, slow scientific progress, concentrate power in a few players, and limit benefits for everyone, including the Global South. An emerging issue tied to this is agentic AI and autonomous systems - models that can act, plan, and interact with the real world over long horizons. Current themes touch on oversight and safety, but they don't yet address new questions around real-time accountability, multi-agent interactions, or how governance must evolve when AI moves beyond chatbots into tools that make decisions or control infrastructure. Another subtle gap is the pace of capability advancement versus governance timelines. AI progress continues to outrun traditional policy cycles; without explicit focus on adaptive, lightweight mechanisms (like living benchmarks or mutual recognition agreements), dialogues may produce static principles that become outdated quickly. These issues cut across all the listed themes and deserve dedicated attention to ensure governance accelerates discovery rather than hinders it. Addressing them pragmatically would help turn good intentions into outcomes that actually serve humanity.

In my opinion, the governance gaps in the four thematic areas I selected are having a real, day-to-day impact on the United States—especially here in the Pacific Northwest where I live and attend classes in Tulalip, Washington. The biggest challenge is the lack of interoperability of governance approaches. A fragmented landscape—EU AI Act rules clashing with lighter-touch U.S. federal policy, plus varying state laws—creates heavy compliance burdens and uncertainty for companies and institutions across the region. This slows global AI deployment, raises costs, and risks ceding ground to competitors. Safe, secure, and trustworthy AI gaps make it harder to scale frontier systems responsibly. Without shared benchmarks and verifiable testing, we end up with either over-cautious delays or unaddressed risks, while local data-center growth faces energy and community pushback. Transparency, accountability, and human oversight shortfalls erode public trust and create liability headaches, especially as AI agents become more autonomous. On open-source software, open data, and open AI models, national-security export controls create tension: they protect against misuse but limit collaboration, talent flow, and the innovation that has historically powered U.S. leadership. Yet these gaps also open huge opportunities. If we close them pragmatically, the U.S. can set global standards that favor innovation over restriction. Here in Tulalip and the broader Pacific Northwest, that means stronger economic growth, attracting top talent, and turning AI into real breakthroughs in medicine, climate, and education. Open models could democratize access and keep American tech competitive. Addressing these issues thoughtfully would let our country, region, and sector lead responsibly while maximizing AI's benefits for everyone.

In my opinion, the AI Dialogue (Global Dialogue on AI Governance) can play a vital role in advancing international cooperation by serving as a practical, inclusive platform that turns fragmented national efforts into coordinated progress. Established under UN General Assembly Resolution 79/325, it brings governments, industry, civil society, and experts together annually to discuss pressing challenges and share best practices. Its greatest value lies in bridging divides—especially between major AI powers and the Global South—by fostering mutual understanding rather than imposing top-down rules. Specifically, it can: Promote interoperability of governance approaches through agreements on mutual recognition of safety certifications and shared benchmarks for catastrophic risks. This would reduce the costly patchwork of incompatible regulations that currently hinders cross-border AI deployment. Advance safe, secure, and trustworthy AI by anchoring discussions in evidence from the Independent International Scientific Panel on AI, focusing on verifiable testing instead of vague principles. Strengthen transparency, accountability, and human oversight via open exchanges on real-world implementations. Support open-source software, open data, and open AI models by encouraging collaborative initiatives that democratize access, accelerate innovation, and prevent power concentration. From my perspective, living and attending classes in Tulalip, Washington, this matters deeply. As someone studying here in the Pacific Northwest—right in the heart of tech innovation—I see how regulatory confusion already affects local opportunities. Successful international cooperation could ease compliance burdens on American companies, attract global talent to our region, and ensure AI breakthroughs in areas like healthcare, climate solutions, and education actually reach communities like mine. It would help the U.S. maintain leadership while making sure benefits are shared more broadly, rather than concentrated among a few players. To succeed, the Dialogue must prioritize concrete deliverables—like joint workstreams with deadlines—over endless talks. If it delivers even modest agreements on interoperability and shared safety tools, it will meaningfully strengthen global cooperation and help AI serve humanity faster and safer.

In my opinion, the Global Dialogue on AI Governance should actively build upon several existing initiatives and mechanisms rather than starting from scratch. Key ones include: The Independent International Scientific Panel on AI (established alongside the Dialogue in Resolution 79/325), which can provide evidence-based assessments on risks and opportunities. The OECD AI Principles and Global Partnership on AI (GPAI), which offer practical guidance on trustworthy AI and multistakeholder collaboration. UNESCO's Recommendation on the Ethics of AI and its policy dialogues, strong on human rights, ethics, and capacity-building for developing countries. G7 Hiroshima AI Process, AI Safety Summits (e.g., Bletchley Declaration), and the ITU's AI for Good summits, which focus on safety testing, real-world applications, and inclusive innovation. Broader UN efforts like the Global Digital Compact and the former High-Level Advisory Body on AI. From my perspective, living and attending classes in Tulalip, Washington, these initiatives already generate valuable knowledge and best practices, but they often remain fragmented or limited in reach—especially for communities and regions outside major tech hubs. The added value the AI Dialogue can bring is its truly global, inclusive UN platform. It can connect these efforts, promote interoperability of governance approaches, facilitate mutual recognition of safety standards, and turn high-level principles into concrete workstreams with deadlines. It could also amplify open-source collaboration and ensure Global South voices help shape capacity-building that actually delivers resources, not just discussion. By focusing on pragmatic outcomes—like shared benchmarks and reduced regulatory friction—the Dialogue can help prevent fragmentation while accelerating safe, beneficial AI that reaches places like my community.

Different stakeholders can contribute meaningfully to the AI Dialogue in complementary ways: Governments should lead on policy alignment, share national best practices, and commit to concrete workstreams like mutual recognition of safety standards. Industry and tech companies can provide real-world technical insights, data on capabilities and risks, and examples of responsible deployment and open-source efforts. Civil society and academia bring ethical perspectives, human rights analysis, and independent research—especially important for transparency and accountability. Global South representatives and youth ensure inclusive voices shape outcomes, particularly on capacity-building and equitable access. From my perspective, living and attending classes in Tulalip, Washington, I believe the Dialogue should actively invite submissions and participation from students, regional communities, and smaller organizations so that AI governance reflects everyday realities, not just headquarters views. Recommendations for format and structure: The Dialogue should keep its annual, two-day format (alternating between Geneva and New York) but make sessions more action-oriented. Structure each meeting with: A high-level governmental segment for commitments, Presentation of the Independent International Scientific Panel's report, Interactive multistakeholder thematic roundtables focused on my priority areas (safe/trustworthy AI, interoperability, transparency/oversight, and open-source), Dedicated working groups with clear deadlines for deliverables like shared benchmarks. Prioritize interactive formats over speeches—breakout sessions, solution labs, and public submission portals—so discussions lead to tangible outcomes rather than repetition. Strong support for developing country participation through voluntary funding will make it truly global. This approach would help the Dialogue move from talk to real international cooperation that benefits communities like mine.

In my opinion, several important voices remain underrepresented in global AI governance discussions, despite efforts to make them more inclusive. Key underrepresented groups include: Communities from the Global South (especially smaller developing countries, Sub-Saharan Africa, Southeast Asia, and Small Island Developing States), whose perspectives on infrastructure gaps, bias amplification, and local impacts are often sidelined by Global North priorities. Indigenous Peoples and Tribal Nations, whose knowledge systems, data sovereignty concerns, cultural preservation needs, and rights to self-determination are frequently overlooked. Women and gender minorities, who make up a small share of the AI workforce and whose experiences with bias in datasets and applications are under-addressed. Youth, students, local communities, educators, and everyday users (including underpaid data workers), whose lived experiences with AI in schools, workplaces, healthcare, and environments rarely shape high-level agendas. From my perspective, living and attending classes in Tulalip, Washington, I see how these gaps matter close to home. Indigenous communities like mine bring unique insights on data sovereignty and culturally grounded AI, yet they are rarely at the center of global talks. How to include them more effectively: Provide dedicated funding, travel support, and translation services for Global South and Indigenous participants. Create structured mechanisms such as regional pre-consultations, youth advisory panels, and Indigenous co-creation sessions. Use public submission portals, citizen assemblies, and hybrid formats that lower barriers for students and civil society. Ensure the Independent International Scientific Panel and working groups actively recruit diverse experts, including social scientists and community representatives, rather than defaulting to "usual suspects." Explore innovative tools like AI proxies for hard-to-reach voices, while always prioritizing real human participation and free, prior, and informed consent. Greater inclusion would make governance more legitimate, practical, and equitable—helping AI benefit communities like mine instead of deepening divides. The AI Dialogue has a strong opportunity to lead on this by turning inclusion commitments into real structural changes.

In my opinion, the AI Dialogue could foster far more meaningful engagement by moving beyond traditional speeches to dynamic, interactive formats that encourage real collaboration. Innovative formats I recommend include: Solution Labs and Design Sprints: Small, mixed groups (governments, industry, civil society, students, and Indigenous voices) tackle specific challenges—like shared safety benchmarks or interoperability agreements—in focused, time-bound workshops. This turns discussion into tangible prototypes or commitments. Citizen and Youth Deliberative Panels: Draw from successful models like global citizen assemblies. Invite randomly selected or regionally representative participants (including youth and underrepresented communities) for structured dialogues that feed directly into plenary sessions, grounding high-level talks in lived experiences. Informal "Off-the-Record" Learning Cafés and Speed Networking: Create safe spaces for diplomats, technologists, and community representatives to ask questions, share failures, and build relationships without political pressure—proven effective in Global Digital Compact preparations. Hybrid Digital Tools and Interactive Dashboards: Use real-time polling, AI-facilitated breakout rooms, and public submission platforms so remote participants (especially from the Global South or places like my community in Tulalip) can contribute equally. Live demos of open-source models or risk benchmarks could make abstract topics concrete. From my perspective, living and attending classes in Tulalip, Washington, these formats would help ensure diverse voices—including Indigenous and student perspectives—aren't just in the room but actively shaping outcomes. Pairing them with clear deadlines for working groups would keep energy high and deliver concrete progress rather than repetition. By prioritizing action-oriented, inclusive engagement, the Dialogue can become a genuine driver of pragmatic international cooperation on safe, open, and interoperable AI.

In my opinion, several existing policies, practices, platforms, and approaches already offer concrete solutions to AI governance challenges, particularly in the areas I prioritize: safe/trustworthy AI, interoperability, transparency/oversight, and open-source models. Notable examples include: The NIST AI Risk Management Framework (AI RMF) and ISO/IEC 42001 provide practical, voluntary guidance for risk assessment, transparency, and lifecycle oversight. They emphasize verifiable testing and human-in-the-loop controls without heavy regulation. The EU AI Act demonstrates a risk-based approach with clear obligations for high-risk and general-purpose AI systems, including transparency requirements and conformity assessments-though it highlights the need for better global interoperability. OECD AI Principles and the G7 Hiroshima AI Process promote shared international standards, codes of conduct, and mutual recognition, helping reduce regulatory fragmentation. Open-source initiatives like Hugging Face's responsible AI tools, FINOS Common Controls for AI Services, and public-private benchmark collaborations offer transparent model scrutiny, safety testing protocols, and community-driven validation that accelerate safe innovation. Enterprise platforms such as Credo AI and Fiddler AI enable automated monitoring, bias detection, and compliance tracking across the AI lifecycle. From my perspective, living and attending classes in Tulalip, Washington, these examples show that effective governance works best when it focuses on evidence-based benchmarks, open collaboration, and pragmatic interoperability rather than one-size-fits-all rules. They help prevent power concentration while supporting breakthroughs that could benefit communities like mine in healthcare, education, and environmental solutions. The AI Dialogue could build on these by scaling shared benchmarks, promoting mutual recognition agreements, and creating inclusive workstreams that turn principles into actionable, adaptable outcomes.