Permanent Mission of Japan to the UN/Government of Japan

a. Overall structure We highly value the Co-Chairs' strong commitment to a genuinely multi-stakeholder approach in both the preparatory process and the Global Dialogue itself. To steer the development and use of AI in the right direction, it is essential to reflect a balanced range of knowledge, technologies, and perspectives of diverse actors from across the AI ecosystem, going beyond governmental bodies. Industry holds much of the technology. Cutting-edge expertise resides in academia. ITU, UNESCO, and OECD have been leading international discussions on AI governance. It is, therefore, imperative that the Global Dialogue ensure meaningful engagement by these multi-stakeholders through clear institutional arrangements and practical modalities. From this perspective, Japan welcomes that most of the two-days of meetings of the Global Dialogue, including its thematic sessions, will follow a multi-stakeholder format. In the same vein, a session dedicated to feeding recommendations and issues distilled in the thematic sessions back to the plenary will be a good opportunity to ensure transparency and inclusiveness for the sake of a truly effective and legitimate multi-stakeholder approach. The thematic sessions will also allow Member States to deepen interactive discussions with other stakeholders, including members of the Independent Scientific Panel, within a limited two-day timeframe. b. Governmental segment Regarding the proposed "High-level governmental plenary segments" which are to take place in the same modalities in two parts on July 6th, presence of high-level representatives of Member States may require enough time to deliver national statements on the first day of the Global Dialogue. However, the Global Dialogue must be structured to ensure its multi-stakeholder nature, as highlighted above. Further extension of the governmental segments, or dedicating a whole day to them, would go against the innovative nature of the Global Dialogue and its genuinely multi-stakeholder approach. In addition, we recall that Resolution 79/325, in its paragraph 5, decided that the Global Dialogue is "to be composed of a multi-stakeholder plenary meeting with a high-level governmental segment", without "plenary". The title of the sessions in the proposed structure should accurately reflect what was agreed by the Member States.

Safe, secure and trustworthy AI;AI capacity-building

The priority themes for Japan are "Safe, Secure, and Trustworthy AI" and "Capacity Building" since these are two of the most relevant elements toward safe, secure and trustworthy global AI ecosystem. Japan stands ready to contribute to the discussions at the Global Dialogue on AI Governance especially under these two themes. a. Safe, Secure, and Trustworthy AI Under this overarching theme, all stakeholders can also discuss and share good practices on key issues, such as information accuracy, risk mitigation, privacy and data protection, and the prevention of misuse and illegal activities. Japan has been leading international discussions through the Hiroshima AI Process. Today, 68 countries and region, including those from the Global South, endorse this vision to realize safe, secure, and trustworthy AI. The Global Dialogue should aim to foster synergies among various international efforts, including through its dedicated "Dialogue of Dialogues" session, so that the international community can make the most of AI while mitigating its risks. We look forward to fruitful deliberations leading to safe, secure, and trustworthy AI for all. b. Capacity building Capacity building is also critical. The Global Dialogue offers a prime opportunity to showcase good practices on capacity building by multi-stakeholders. For example, Japan, for its part, has been actively engaged in providing institutional development, governance support, human resources development, and capacity-building assistance in India, Africa, ASEAN, and Central Asia. A distinguishing feature of Japan's approach is that it goes beyond mere one-sided assistance and focuses on co-creating AI solutions in a manner tailored to the respective needs of each country. A number of capacity building initiatives by various UN agencies are also underway. Mapping the existing mechanisms and initiatives across the UN system would enable developing countries to make the best use of a variety of in-kind contributions already undertaken based on their respective needs. Such an exercise would also help to identify any gaps between what these existing frameworks can offer and what is actually needed.

See the answer to Question 8