World Trade Organization

• Success at the first Dialogue should be measured by the quality of the shared understanding it produces. The Dialogue could focus on providing a comprehensive surface area on where AI intersects with society at large and implications of the various public policy dimensions including trade, data governance, competition, labour, environment. Given that AI's inputs from data, compute, models, and semiconductors to its outputs are inherently cross-border, trade policy is woven through almost every one of these dimensions. This mapping will help governments and stakeholders understand the variety of existing policy tools and multilateral processes that can be leveraged, and recognise where genuine gaps need to be addressed. This can be captured in the proposed Chair's Summary. The Dialogue should also affirm inclusiveness as a cross-cutting objective, and recognise that trade has a specific contribution to make alongside other policies. Open and predictable trade rules can lower the cost of AI-related goods and services and widen access
• balanced IP rules can support both innovation and diffusion
• international standards work can curb regulatory fragmentation
• and trade-related capacity-building can help developing economies build the foundations they need to participate in AI-driven trade. In this regard, the Dialogue could usefully identify a few concrete priorities for cooperation — particularly on access to compute, quality data, and skills — and signal openness to new cooperation models. Finally, the Dialogue should reaffirm that meaningful global norms should be built through multilateral cooperation to avoid fragmentation, and recognise the complementary role that multilateral organizations, including the WTO, can play in delivering them.

• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI

The WTO Secretariat believes that AI will be transformative to trade. Our simulations suggest that global trade could rise by nearly 40% by 2040. Global GDP could increase by 12-13%. Our simulations show that narrowing the digital and AI divides could almost double the GDP gains for low-income economies from 8 to 15%. Under the right conditions, AI can enable more inclusive growth through trade. It has the potential to help developing economies to leapfrog and enable small businesses to participate in global trade. On the other hand, the uneven diffusion of AI technologies, persistent skills gaps and limited digital infrastructure risk reinforcing existing disparities, both between and within economies. From the inputs required to develop AI, to compute, data and ultimate AI applications, the AI value chain is inherently cross-border. That is why trade policy is central - it directly shapes the availability, affordability, and diffusion of AI and the goods and services that make it possible. Managing the transition to an AI driven economy does not only require strong domestic policies such as investment in digital infrastructure and retraining programs; it also requires ambitious trade policies to enhance access to key inputs, promote services that underpin AI development and diffusion, and foster cross-border data flows. There is growing interest from members in having more conversations on emerging issues like AI at the WTO. Members have started to submit notifications on AI-related measures as well as specific trade concerns about AI at the WTO. Hence, we feel that it is important that trade be considered as part of the Dialogue. This could be incorporated as part of the discussions on AI opportunities and implications and/or bridging the AI divides.