WTO World Trade Report 2025: How AI Is Reshaping Global Trade

Understanding the WTO World Trade Report 2025

The WTO World Trade Report 2025 arrives at a critical inflection point for the global economy. Titled “AI and Trade,” this comprehensive report examines how artificial intelligence is fundamentally transforming international commerce — from the way goods cross borders to how services are delivered, regulated, and consumed worldwide. As WTO Director-General Ngozi Okonjo-Iweala stated in the report’s foreword, “AI can become a powerful driver of inclusive, trade-led growth… but only if economies invest in the right enabling policies and cooperate to prevent fragmentation of the digital economy.”

The report’s central thesis is both promising and cautionary. AI technologies offer extraordinary potential to reduce trade costs, expand market access for small and medium enterprises, and make previously non-tradable services accessible across borders. Yet these gains are not automatic. Without deliberate investment in digital infrastructure, workforce skills, and coordinated international regulatory frameworks, the AI revolution risks deepening existing economic divides rather than closing them. For a broader perspective on how global trade is evolving, explore our analysis of the WTO Global Trade Outlook 2025, which provides essential context on trade volume trends and forecasts.

What makes the 2025 edition particularly significant is its scope. Drawing on original WTO economic simulations, a joint WTO–ICC business survey of firms across income groups, and extensive analysis of trade policy landscapes in over 140 economies, the report builds the most comprehensive evidence base yet assembled on the AI-trade nexus. It spans goods, services, data flows, intellectual property, competition policy, energy infrastructure, and workforce development — recognizing that AI’s impact on trade cannot be understood through any single policy lens.

AI as a Catalyst for Trade-Led Growth

The headline projections from the WTO’s economic modeling paint a compelling picture. Under various AI adoption scenarios, global trade is projected to rise by 34–37% by 2040, with global real GDP increasing by approximately 12–13% over the same period. These are not hypothetical figures but carefully modeled outcomes based on different assumptions about AI diffusion rates, infrastructure investment, and policy environments across economies.

Perhaps the most striking finding concerns digitally deliverable services. The report projects that trade in these services — encompassing everything from financial analytics and legal research to software development, design, and consulting — could grow by approximately 42% by 2040. This reflects AI’s unique ability to make services that were traditionally delivered face-to-face or within national borders suddenly tradable across continents. A consulting firm in Nairobi can now deliver AI-powered market analysis to clients in London. A medical imaging startup in Mumbai can process diagnostic scans for hospitals in São Paulo.

The WTO–ICC business survey reinforces these projections with ground-level evidence. Among firms currently using AI, approximately 90% report tangible trade-related benefits, while 56% cite improved ability to manage trade risks — including compliance monitoring, currency fluctuation analysis, and supply chain disruption forecasting. These are not marginal improvements but transformative capabilities that reshape how firms engage with international markets.

However, the report sounds an important note of caution. AI adoption remains unevenly distributed. The survey found that 41% of small firms use AI compared to over 60% of large firms, and in low and lower-middle income economies, fewer than one-third of firms report any AI use. This adoption gap translates directly into competitive disadvantage, creating a self-reinforcing cycle where the firms and economies most in need of productivity gains are least able to access them.

How AI Lowers Trade Costs and Boosts Productivity

One of the report’s most valuable contributions is its detailed analysis of the mechanisms through which AI reduces trade costs. These mechanisms operate at multiple levels — from individual firm operations to national customs systems to global supply chain coordination.

At the firm level, AI streamlines export documentation, automates regulatory compliance checks, and optimizes logistics routing. Machine learning algorithms can predict customs delays, identify the most cost-effective shipping routes, and flag potential compliance issues before goods reach the border. For MSMEs that lack dedicated trade compliance departments, these capabilities are particularly transformative — effectively giving a ten-person company the trade management sophistication previously available only to multinational corporations.

At the national level, AI-powered customs systems can dramatically reduce clearance times. The report documents how AI-enhanced risk assessment tools allow customs authorities to focus inspections on genuinely high-risk shipments rather than conducting random or universal checks. This reduces wait times, lowers costs for compliant traders, and improves the detection of illicit goods — a rare win-win-win in trade policy.

The productivity effects extend beyond cost reduction. AI enables firms to identify and enter new markets by analyzing demand patterns, regulatory requirements, and competitive landscapes across countries. Natural language processing tools help firms navigate foreign-language regulations and customer communications. Computer vision systems enable automated quality control that meets diverse international standards. Each of these capabilities makes it easier and cheaper for firms to trade across borders, expanding the universe of commercially viable international transactions.

The $2.3 Trillion AI-Enabling Goods Market

The report introduces a critical concept that reshapes how we should think about AI and trade: the AI-enabling goods market. This encompasses the physical infrastructure of AI — from raw materials like rare earth elements and specialty chemicals to semiconductors, high-performance computing equipment, networking hardware, and data storage systems. In 2023, global trade in these AI-enabling goods reached approximately US$2.3 trillion, making it one of the largest and most strategically significant segments of international commerce.

This figure underscores a fundamental reality: AI is not purely a software phenomenon. Every AI system depends on a vast, globally distributed supply chain of physical inputs. Training a large language model requires thousands of specialized GPUs manufactured from materials sourced across multiple continents, assembled in specialized fabrication facilities, and deployed in data centers that consume enormous quantities of electricity. The report notes that data centers already consume approximately 1.5% of global electricity — a figure projected to grow substantially as AI workloads intensify.

The trade policy implications are profound. Tariff barriers on AI-enabling goods directly affect the affordability and accessibility of AI technologies. The report reveals that some low-income economies maintain bound tariffs on AI-enabling goods as high as 45%, effectively pricing their firms and researchers out of the global AI ecosystem. This finding strengthens the case for broader participation in agreements like the WTO Information Technology Agreement (ITA), which eliminates tariffs on a wide range of technology products.

The concentration of AI hardware production also raises strategic concerns. A small number of economies and firms dominate the manufacture of advanced semiconductors and AI accelerators. The report documents how subsidies targeting AI-related products have surged since 2010, exceeding 15% of all industrial subsidies at their peak — with over 98% of those measures originating in high and upper-middle-income economies. This concentration creates both supply chain vulnerabilities and competitive imbalances that trade policy must address.

Digital Trade Frameworks and Cross-Border Data Flows

Data is the fuel of artificial intelligence, and the WTO report dedicates substantial analysis to the trade dimensions of data governance. Cross-border data flows are essential for AI development and deployment — training data must be collected, processed, and shared across jurisdictions; AI services must be delivered internationally; and the models themselves rely on continuous access to fresh information from global sources.

Yet the regulatory landscape for cross-border data flows remains fragmented and often contradictory. The report documents a growing patchwork of data localization requirements, transfer restrictions, and privacy regulations that collectively raise costs and constrain AI innovation. While many of these measures serve legitimate policy objectives — protecting privacy, ensuring national security, or supporting domestic industry development — their cumulative effect can be to segment the global data economy into isolated pools, reducing the quality and availability of training data for AI systems.

The WTO report calls for “proportional data governance that balances privacy, intellectual property, and trade-enabling flows.” This means designing regulations that achieve their protective objectives without imposing unnecessary barriers to legitimate cross-border data use. The report highlights existing WTO transparency tools, such as the ePing alert system, as practical mechanisms for increasing visibility into new data-related trade measures and reducing the surprise factor that currently complicates business planning.

For organizations seeking to understand how digital economy regulations are shaping global commerce, our interactive analysis of the UNCTAD Digital Economy Report 2025 offers complementary perspectives on data governance frameworks and their economic implications.

Bridging the AI Divide: Developing Countries and Inclusive Growth

The most consequential section of the WTO report concerns the distribution of AI’s benefits across different levels of economic development. The analysis presents multiple scenarios that illustrate a stark choice facing the global community: AI can become either a force for economic convergence or a driver of deeper divergence, depending entirely on policy decisions made in the coming years.

In the baseline scenario — where low-income economies do not significantly close their digital infrastructure gap — the projections are sobering. By 2040, high-income economies would see income gains of approximately 14%, middle-income economies 11%, and low-income economies only 8%. This gap represents not just a missed opportunity but a widening of existing inequalities, as the countries most in need of productivity growth fall further behind.

However, the report’s convergence scenarios offer genuine cause for optimism. If developing economies close their digital infrastructure gap by 50% and adopt AI more broadly across their economies, income gains for low-income countries could reach 11% — nearly matching middle-income outcomes. In the most ambitious scenario, combining infrastructure investment with broad AI adoption and supportive trade policies, low-income economies could achieve 15% income growth, actually surpassing the baseline gains for high-income countries.

The practical pathways for developing country participation in the AI economy are specific and actionable. The report identifies several immediate opportunities: adoption and adaptation of pre-trained AI models (rather than building from scratch), data collection and annotation services (where labor cost advantages apply), cloud and hosting services, and niche export services where AI augments existing capabilities. Programs like the WTO’s Aid for Trade initiative, Digital Trade for Africa, and the Women Exporters in the Digital Economy (WEIDE) Fund represent concrete mechanisms for supporting this transition.

Domestic Policy Levers for AI and Trade

The WTO report provides an unusually detailed analysis of how domestic policy choices shape the AI-trade relationship. Rather than treating trade policy in isolation, it examines a comprehensive ecosystem of policies that collectively determine whether a country’s firms and workers benefit from AI-driven trade transformation.

Trade policy itself is the first lever. Tariff levels on AI-enabling goods, services trade restrictions, and market access commitments under the General Agreement on Trade in Services (GATS) all directly affect AI affordability and diffusion. The report argues for updated and broader services commitments that reflect the AI-driven transformation of service delivery, noting that many existing GATS schedules were negotiated before AI capabilities existed.

Intellectual property policy has seen particularly rapid evolution. The number of economies adopting at least one AI-related IP policy grew from 41 in 2017 to 140 in 2024 — a remarkable acceleration that reflects both the urgency of the issue and the diversity of approaches being tried. Key questions include patentability of AI-generated inventions, copyright protection for AI training data, and trade secret frameworks for proprietary models and algorithms. The report calls for coherent IP frameworks that encourage innovation while preventing excessive market concentration.

Competition policy is another critical dimension. Since the launch of ChatGPT in November 2022, the report documents 44 competition measures targeting AI, with over 80% originating in high and upper-middle-income economies. These measures address market power concerns ranging from cloud computing dominance to AI model marketplace concentration to preferential data access. For insights into how governance frameworks are being developed, see our analysis of the ICC AI Governance Standards 2025.

Energy policy emerges as a perhaps surprising but critical factor. With data centers consuming approximately 1.5% of global electricity and AI workloads driving rapid growth in computing demand, access to affordable and reliable energy directly affects where AI infrastructure can be economically deployed. The report reveals a stark disparity: high-income economies account for 69% of global renewable energy policies, while low-income economies account for just 1.5%. This energy policy gap translates into an AI infrastructure gap, as firms and cloud providers preferentially locate data centers in jurisdictions with reliable, affordable power.

Education and workforce development complete the policy ecosystem. The report emphasizes STEM education, AI-specific curricula, reskilling programs, and lifelong learning initiatives as essential complements to trade liberalization. Without a workforce capable of using, adapting, and developing AI tools, market access alone cannot deliver trade benefits. Targeted programs for women entrepreneurs and MSME support are highlighted as particularly high-impact interventions.

International Cooperation and the Role of the WTO

The final major section of the report examines the role of multilateral cooperation in making AI and trade work for all economies. The WTO positions itself as a central — though not exclusive — forum for managing the trade dimensions of AI, while recognizing the need for coordination with other international organizations including WIPO, the OECD, ITU, UN agencies, the World Bank, and the International Trade Centre.

The report identifies several specific roles for the WTO. First, as a custodian of existing trade agreements that remain highly relevant to AI — including the Information Technology Agreement, GATS, the Agreement on Technical Barriers to Trade (TBT), and TRIPS. These agreements provide foundational rules for market access, transparency, and intellectual property that directly affect AI trade, even though they were negotiated before AI became commercially significant.

Second, the WTO can serve as a transparency and surveillance platform. Its committee structures, notification requirements, and monitoring tools provide mechanisms for tracking AI-related trade measures, sharing best practices, and identifying emerging barriers before they become entrenched. The report calls for enhanced use of these existing institutional capabilities rather than the creation of entirely new frameworks.

Third, the WTO’s technical assistance and capacity-building programs offer practical channels for supporting developing country participation in AI trade. Programs like Digital Trade for Africa, Digital Trade for Latin America, and the WEIDE Fund are already operational and can be expanded to incorporate AI-specific components.

The report also addresses the challenge of regulatory interoperability — the growing risk that divergent national AI regulations will create compliance burdens that effectively exclude smaller firms and developing country exporters from AI-intensive trade. It advocates for international standards, mutual recognition mechanisms, and regulatory dialogue to reduce fragmentation without imposing one-size-fits-all approaches. For broader context on how economic institutions are addressing these challenges, explore our coverage of the BIS Annual Economic Report 2025.

Strategic Takeaways for Businesses and Policymakers

The WTO World Trade Report 2025 delivers a message that is simultaneously optimistic and urgent. The economic potential of AI-driven trade transformation is substantial — trillions of dollars in additional trade, significant productivity gains, and the possibility of more inclusive growth that narrows rather than widens the gap between rich and poor economies. But realizing this potential requires coordinated action across multiple policy domains and levels of governance.

For businesses, the immediate implications are clear. Firms that invest in AI capabilities for trade operations — from compliance automation to market intelligence to supply chain optimization — will gain significant competitive advantages in the coming decade. The WTO–ICC survey data suggests this is already happening, with AI-adopting firms reporting measurable improvements in trade performance. Small and medium enterprises should prioritize accessing AI tools through cloud-based services and pre-trained models rather than attempting to build proprietary systems from scratch.

For policymakers, the report provides a comprehensive framework for action. The most effective approaches will be those that address the entire policy ecosystem simultaneously — trade liberalization paired with infrastructure investment, data governance aligned with innovation incentives, workforce development coordinated with market opening. Isolated policy interventions are unlikely to capture the full potential of AI-driven trade transformation.

The report’s convergence scenarios offer a powerful argument for investment in developing country digital infrastructure and AI capability. The difference between the baseline and convergence outcomes for low-income economies — 8% versus 15% income growth by 2040 — represents hundreds of billions of dollars in foregone development if the infrastructure and policy gaps are not addressed. This makes the case for expanded Aid for Trade, technical assistance, and multilateral cooperation not just on equity grounds but on economic efficiency grounds as well.

As Director-General Okonjo-Iweala emphasizes, “Whether AI becomes a force for convergence or for divergence will depend on the choices we make today.” The WTO World Trade Report 2025 provides the analytical foundation and policy roadmap for making the right choices. For organizations looking to explore the original 144-page report in depth, our interactive experience above transforms this essential document into an engaging, navigable format that makes its insights accessible to stakeholders at every level.