Permanent Mission of Viet Nam to the United Nations

1. On overall structure and objectives, the Dialogue should preserve its intergovernmental character while leveraging multi stakeholder expertise. Member States must remain central in setting priorities. Thematics sessions should bring together governments, industry, international organizations, academia and the technical community to exchange concrete practices, technical insights and policy needs. The Dialogue should be explicitly results oriented, aiming at policy recommendations, cooperative initiatives and actionable roadmaps with measurable follow up. 2. Bridging AI divides requires targeted focus on AI relevant infrastructure and computational capacity, including secure connectivity, data centers, cloud and compute resources, and mechanisms for more equitable access, particularly for developing countries. Key areas for discussion and action include scaling infrastructure in developing countries, promoting mechanism that enable access to shared infrastructure, ensuring that infrastructure expansion aligns with cybersecurity and data protection standards. 3. The Dialogue should encourage responsible development, use and, where appropriate, sharing of data resources to ensure that AI systems reflect diverse social contexts, languages and cultural identities, and to mitigate data bias and discrimination. Priority areas include data quality and representativeness, privacy and ethical safeguards, inclusion of under represented languages and communities, and governance frameworks that enable data to be used for the public good. 4. The Dialogue could also play a useful role in promoting convergence around high level technical requirements for AI systems. These should cover, inter alia quality of service parameters such as latency, reliability and robustness; safety and resilience; cybersecurity. Advancing shared understandings in these areas will help ensure stability, security and trustworthiness in AI deployment, and support a more coherent and cooperative international AI governance landscape.

• AI capacity-building
• Transparency, accountability, and human oversight
• Interoperability of governance approaches
• Social, economic, ethical, cultural, linguistic and technical implications of AI

1. On overall structure and objectives, the Dialogue should preserve its intergovernmental character while leveraging multi stakeholder expertise. Member States must remain central in setting priorities. Thematics sessions should bring together governments, industry, international organizations, academia and the technical community to exchange concrete practices, technical insights and policy needs. The Dialogue should be explicitly results oriented, aiming at policy recommendations, cooperative initiatives and actionable roadmaps with measurable follow up. 2. Bridging AI divides requires targeted focus on AI relevant infrastructure and computational capacity, including secure connectivity, data centers, cloud and compute resources, and mechanisms for more equitable access, particularly for developing countries. Key areas for discussion and action include scaling infrastructure in developing countries, promoting mechanism that enable access to shared infrastructure, ensuring that infrastructure expansion aligns with cybersecurity and data protection standards. 3. The Dialogue should encourage responsible development, use and, where appropriate, sharing of data resources to ensure that AI systems reflect diverse social contexts, languages and cultural identities, and to mitigate data bias and discrimination. Priority areas include data quality and representativeness, privacy and ethical safeguards, inclusion of under represented languages and communities, and governance frameworks that enable data to be used for the public good. 4. The Dialogue could also play a useful role in promoting convergence around high level technical requirements for AI systems. These should cover, inter alia quality of service parameters such as latency, reliability and robustness; safety and resilience; cybersecurity. Advancing shared understandings in these areas will help ensure stability, security and trustworthiness in AI deployment, and support a more coherent and cooperative international AI governance landscape.