xidian university

Key markers of success include reaching a ministerial-level political declaration and actionable framework that embeds vision of people-centered, security-development balanced AI governance, while upholding the principles of AI equality. It also requires establishing a science-based global AI risk assessment and early warning system, building mutually recognized cross-border governance rules to break down fragmented barriers, and launching a capacity-building fund and network to bridge the digital divide for Global South countries. Additionally, the dialogue should create a sustainable multi-stakeholder platform for regular exchanges and a long-term tracking mechanism with annual progress reports to ensure governance commitments are implemented. Ultimately, success lies in fostering a fair, equitable global AI governance system that amplifies the voices of developing nations and translates consensus into tangible benefits for all humanity.

• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Transparency, accountability, and human oversight
• Open-source software, open data and open AI models

I've chosen these four areas, because they address the dual imperatives of inclusive AI development and responsible AI governance-core tenets of a people-centered, equitable global AI order. First, AI capacity-building is foundational to bridging the digital divide. it empowers developing countries to participate in AI innovation and governance on an equal footing, rather than being marginalized as passive rule-followers. Second, examining the social, economic, ethical, cultural, linguistic and technical implications of AI ensures governance is not limited to technical risk mitigation alone; it safeguards diverse values, protects linguistic diversity, and prevents AI from exacerbating social inequalities or eroding cultural heritage. Third, transparency, accountability, and human oversight can help counter the "black box" problem of AI systems, ensure human agency over algorithmic decisions, and establish clear liability frameworks for AI-related harms. Finally, open-source software, open data and open AI models promote collaborative innovation, reduce barriers to access for resource-constrained regions, and foster a decentralized AI ecosystem that resists monopolies over technology and data. Together, these areas turn abstract governance principles into actionable, inclusive solutions that benefit all stakeholders, especially the Global South.

1. On AI capacity-building, a North-South divide persists. Developed nations monopolize core tech and rule-setting, leaving the Global South reliant on the North for technological advancement. In Asia, fragmented national capacities and weak cross-border coordination hinder inclusive progress, with middle-income states lagging in core innovation despite rapid AI adoption. 2. For social-economic-ethical-cultural implications, Western-centric global norms ignore non-Western values, linguistic diversity, and collective welfare. We should people-centered AI ethics to protect cultural heritage and linguistic equity, but Asia and the Global South in general face AI-driven erosion of local languages, traditional knowledge, and widening urban-rural inequality. 3. Transparency and accountability suffer from a lack of global binding standards and weak regulatory capacity in developing regions. Most Global South nations lack mechanisms to hold foreign tech firms accountable for algorithmic bias or harms, leaving populations vulnerable. 4. Open-source AI gaps, rooted in unequal access to data, computing power, and governance, limit the Global South's ability to customize models for local needs. Most countries in Asia and the Global South at large remain passive users rather than active contributors to global open standards, risking digital marginalization. Overall, these gaps entrench a hierarchical global AI order; targeted, inclusive governance advances are critical to empowering the Global South and building an equitable AI ecosystem.

The UN Global Dialogue on AI Governance can serve as a critical neutral multilateral platform to amplify the voices of the Global South, ensuring diverse perspectives shape equitable global AI rules. It can also catalyze targeted consensus-building on high-priority areas like AI capacity-building, transparency, and open-source collaboration, translating broad principles into actionable frameworks that address the unique needs of developing regions. Most importantly, the Dialogue can lay the foundation for a sustained follow-up mechanism—including joint capacity-building programs and annual progress reviews—to turn discussions into long-term, inclusive cooperation that bridges the North-South AI governance gap.

The AI dialogue should build upon inclusive, development-oriented multilateral initiatives and South-South cooperation mechanisms that center equity and practical capacity-building, while connecting with global frameworks to avoid fragmentation.

The AI Dialogue should ensure that it is fully accessible and inclusive to developing countries by embedding equity into every design and operational stage, starting with removing financial and logistical barriers through a dedicated inclusivity fund to cover participation costs for least developed countries, a hydbrid format and time zone-friendly scheduling that avoids disadvantaging the Global South. It must also guarantee linguistic accessibility and providing real-time interpretation, while centering discussion topics on Global South-specific challenges like agricultural AI and cultural heritage preservation instead of Western-centric agendas. To ensure equitable representation and decision-making power, the UN should enforce quotas for developing country participants (with sub-quotas for women and youth), balance speaking time across stakeholder groups, and prioritize questions from developing nations during sessions. Additionally, capacity-building workshops and mentorship programs can be put into place and empower delegates from developing countries to engage meaningfully. Finally, the UN should avoid allowing Western corporate or multistakeholder groups to overshadow developing country voices, fostering a space where the Global South is a co-shaper of global AI governance rules rather than a passive participant.

The most underrepresented voices are those of developing countries (especially least developed and small island states), indigenous communities, linguistic minorities, and grassroots stakeholders in the Global South. In the current discussion, many perspectives rooted in non-Western values such as collective welfare and cultural heritage preservation are somehow marginalized by Western-dominated discourse. To include these groups, the UN should establish dedicated grants to cover participation and capacity-building costs, host regional pre-dialogue consultations to consolidate their demands, provide multilingual interpretation and design agendas around Global South-specific challenges, set quotas for their representation in steering bodies and reserve speaking time in plenary sessions, and ensure dialogue outcomes explicitly address their priorities like AI capacity-building funds and safeguards against algorithmic discrimination.

The engagement formats should shift the dialogue from a "talk-only" event to a collaborative platform, ensuring that marginalized groups are not just participants but co-designers of global AI governance solutions. For instance, regional, problem-solving workshops can be organized to include regional grassroots stakeholders, policymakers, and technologists to co-design actionable plans to solve probelms, avoiding abstract debates. Alternatively, North-South Pairing Roundtables can be arranged to pair delegations from developed and developing countries in small, facilitated groups to address specific gaps together and also ensure priorities like technology transfer and linguistic equity are not sidelined. These formats can be more problem-driven and collaboration-oriented.

1. China proposed Global AI Governance Initiative that advocates a multi-stakeholder, inclusive global governance system that respects national sovereignty, bridges the digital divide, and supports developing countries in AI capacity-building; promotes sharing of open-source models, technical standards, and governance experience with Global South nations. 2. AI is used to enhance productivity and sustainability in rural China's agricultural sector. Smart breeding, for example, is powered by AI, unlocking potential to develop high-yield, climate-resilient "super crops" in China. China also establishes joint research labs together with ASEAN and Africa to co-develop region-specific AI solutions, sharing expertise in AI model training, big data processing, and climate-adaptive agricultural AI.