AI in the Chain

Introduction

Global trade is often portrayed as fragile and precarious. Headlines about tariffs, wars and supply‑chain disruptions create the impression that cross‑border commerce is in retreat. Yet the 2025 edition of the DHL Trade Atlas paints a very different picture: trade is forecast to continue growing over the next five years, with growth slightly outpacing global GDP. Even under severe scenarios—such as a second Trump administration imposing all proposed tariff increases and trading partners retaliating—global goods trade is expected to grow, albeit more slowly. The report offers vital clues about which countries, regions and sectors will drive this expansion and how geopolitical turbulence might reshape flows. For supply‑chain leaders and AI practitioners, these insights are a treasure trove: combining predictive analytics and generative AI with up‑to‑date trade intelligence can unlock new opportunities and enhance resilience.

This article distils the ten key takeaways from the DHL report and connects them to AI‑driven supply‑chain strategies. It concludes with a hands‑on exercise showing how generative AI can help analyse trade growth scenarios.

Key Findings from the DHL Trade Atlas

1. Faster growth amid greater uncertainty

The DHL report aggregates forecasts from four respected institutions and predicts that global goods trade will grow modestly faster between 2024 and 2029 than it did in the previous decade. This baseline assumes that only some of the proposed U.S. tariff increases will materialise; under those conditions, trade growth is expected to match or slightly outpace global GDP. The drivers of this uptick include expanding cross‑border e‑commerce, rising consumption in emerging markets and companies reconfiguring supply chains to diversify risk. Yet the outlook is clouded by record‑high uncertainty: aggressive tariff threats, heightened geopolitical tensions and economic headwinds could easily slow the pace. The report notes that if all proposed U.S. tariffs were implemented and other countries retaliated in turn, global trade would still grow but roughly one‑quarter slower. History reminds us that trade has absorbed shocks before—the 2008 financial crisis and the COVID‑19 pandemic caused steep but short‑lived drops in trade volumes. AI implication: supply‑chain planners should adopt probabilistic forecasting and scenario analysis rather than relying on a single baseline. Machine‑learning models can ingest macro indicators, policy announcements and sentiment data to generate distributions of possible outcomes. Generative AI can then craft narrative explanations for executives and board members, translating complex scenarios into clear action items. An AI‑enabled dashboard might, for example, notify planners when new tariff legislation increases downside risk and automatically adjust safety stock levels or sourcing plans.

2. Tariff threats will dampen, not derail, trade

Even if all tariff increases proposed by a second Trump administration were enacted and partners retaliated, the DHL report concludes that global trade growth would remain positive. This finding underscores the resilience of international commerce: trade barriers can slow momentum but rarely halt it. Historical episodes confirm this pattern; crises such as the 2008 financial crash, the escalation of U.S.–China trade tensions, the COVID‑19 pandemic and wars in Ukraine and Gaza caused temporary downturns yet did not lead to sustained contraction. Trade’s durability stems from its role in fostering prosperity and poverty reduction—businesses and governments have strong incentives to keep goods flowing even amid protectionist rhetoric. Nevertheless, tariffs do reshape supply‑chain economics by raising costs, altering comparative advantages and prompting firms to reconsider their supplier mix. AI implication: companies should leverage digital twins and reinforcement‑learning agents to prepare for various tariff scenarios. A digital twin of the supply network can model how a 10 % tariff on Chinese electronics would influence lead times, landed costs and inventory levels. Reinforcement‑learning agents can learn policies for switching suppliers or rerouting shipments when tariffs cross certain thresholds, automatically recommending optimal supplier portfolios. Generative AI can assist by drafting alternative contract terms or supplier communication templates to implement contingency plans quickly.

3. The U.S. still relies on China, indirectly

U.S. direct imports from China have fallen as companies diversify sourcing, but the report warns that made‑in‑China content continues to enter the United States via other countries. Inputs from China are embedded in products assembled in Vietnam, Mexico and other third countries before being shipped to the U.S., meaning that American consumers remain indirectly reliant on Chinese supply chains. The report even cautions that U.S. direct imports from China may be underreported due to trans‑shipment or classification errors. This hidden reliance creates blind spots: a firm may think it has weaned itself off Chinese suppliers only to find that its contract manufacturer in Mexico imports semiconductors from Shenzhen. AI implication: risk‑management tools must map multi‑tier dependencies. Graph neural networks and knowledge‑graph techniques can trace relationships across tiers, linking part numbers, supplier names and country of origin to identify hidden exposures. Natural‑language processing can extract supplier information from invoices and bills of materials, while anomaly‑detection models flag unusual patterns in customs data. Once dependencies are identified, companies can decide whether to dual‑source, invest in domestic production or carry more buffer inventory. A practical example is using a graph‑analysis tool to visualise which assembly plants in Mexico rely on Chinese components, then evaluating alternative sources in Malaysia or South Korea.

4. Geopolitical shifts remain limited

Although tensions between major powers have escalated, the DHL report finds little evidence that the world is splitting into disconnected blocs. Trade between geopolitical blocs declined in 2022 and 2023 but stabilised in 2024. Studies cited in the report warn that a complete split of world trade between rival blocs could cut global GDP by up to 7 %, giving governments strong incentives to maintain openness. Recent policy developments also signal ongoing integration: the EU and Mercosur took a major step toward establishing a free‑trade zone in December 2024, and the U.K. joined the CPTPP shortly thereafter. While near‑shoring and friend‑shoring are part of many corporate strategies, the evidence suggests that global sourcing will remain important. AI implication: supply‑chain organisations should not rush to regionalise everything. AI‑driven network‑design tools can evaluate trade‑offs between near‑shoring (shorter lead times, lower geopolitical risk) and global sourcing (cost advantages, access to innovation). By simulating scenarios such as a 10 % increase in cross‑bloc tariffs or a sudden closure of a strategic shipping lane, digital twins can help planners design resilient yet efficient networks. Generative AI can draft scenario narratives for executives, highlighting trade‑offs, while predictive models monitor early warning signals like changes in UN voting patterns or rising conflict incidents.

5. Past growth leaders: UAE, Vietnam and Ireland

The only countries ranked in the top 30 worldwide for both the speed (growth rate) and scale (absolute amount) of their trade expansion over the past five years were the United Arab Emirates (UAE), Vietnam and Ireland. Each achieved success through different strategies. The UAE leveraged its strategic location at the crossroads of Asia, Europe and Africa, pouring billions into logistics infrastructure (ports, airports and free‑trade zones) and promoting trade‑friendly policies. Vietnam transformed itself into a manufacturing powerhouse, attracting foreign investment in electronics, textiles and footwear, and signing trade agreements with the EU, Japan and other partners. Ireland’s growth reflects its role as a hub for high‑value goods such as pharmaceuticals, medical devices and ICT hardware, aided by a favourable tax environment and skilled workforce. AI implication: firms can use machine learning to analyse trade‑growth “signals” and identify rising hubs for sourcing or distribution. Time‑series models can detect accelerating export volumes in specific industries, while clustering algorithms can group countries with similar trade profiles to spot potential “next Vietnams.” Generative AI can also draft market‑entry strategies tailored to new hubs, including regulatory overviews, logistics partners and workforce considerations. When combined with human expertise, these tools accelerate strategic planning and de‑risk expansion.

6. Long‑distance trade is not dead

Contrary to predictions of widespread localisation, the average distance a traded good travelled during the first nine months of 2024 reached a record 5,000 km. Only 51 % of trade was intra‑regional—the lowest share on record. Several forces drive this: companies continue to source from the most cost‑effective locations, consumers expect variety regardless of origin, and cross‑border e‑commerce platforms have broadened access to distant markets. In addition, long‑haul shipments of commodities and intermediate goods remain essential to global manufacturing networks. The report suggests that disruptions such as the Suez Canal blockage or drought‑induced closures of the Panama Canal remind us how vulnerable these long routes can be to natural and geopolitical shocks. AI implication: route‑optimisation algorithms must account for longer haul distances, geopolitical choke points and carbon‑emission penalties. Reinforcement‑learning models can explore alternative routes when canals close or conflicts erupt, balancing cost, time and sustainability. For example, an AI‑powered system can simulate how diverting ships around the Cape of Good Hope affects fuel costs and lead times for Asia‑Europe trade, or whether shifting some cargo to rail and truck via inland corridors reduces exposure to maritime chokepoints. Including carbon calculations in these models enables firms to align route decisions with sustainability targets and regulatory requirements.

7. Future growth leaders: India, Vietnam, Indonesia and the Philippines

Looking ahead, India, Vietnam, Indonesia and the Philippines are forecast to rank among the top 30 countries for both speed and scale of trade growth. India stands out as the country with the third‑largest expected increase in absolute trade, generating roughly 6 % of additional global trade over the next five years, behind only China (12 %) and the United States (10 %). These economies share demographic advantages (large, youthful populations) and policy reforms that attract foreign direct investment. India and Indonesia are investing heavily in ports, highways and digital infrastructure to cut logistics costs and improve reliability. Vietnam continues to benefit from its role as an alternative to China for electronics and apparel manufacturing, while the Philippines leverages its English‑speaking workforce and growing business‑process outsourcing sector. Capacity constraints, regulatory bottlenecks and political risks could, however, introduce volatility. AI implication: companies should monitor port capacity, customs regulations and infrastructure development in these countries. Predictive models can estimate lead‑time variability and flag when to diversify or invest. For example, an AI system might track dwell times at major Indian ports, correlate them with monsoon patterns or labour strikes and recommend pre‑booking additional vessel slots during high‑risk periods. Generative AI can produce localised compliance checklists, language translations and training materials for new suppliers. Digital twins can test how adding distribution centres in Chennai or Jakarta would affect service levels across Asia.

8. Standout regions: South Asia, Sub‑Saharan Africa and Southeast Asia

South Asia, Sub‑Saharan Africa and Southeast Asia are projected to achieve compound annual trade‑growth rates of 5–6 % between 2024 and 2029, far outpacing other regions. This reflects rapid population growth, increasing industrialisation and policy reforms such as the African Continental Free Trade Area and ASEAN integration. In absolute terms, however, Europe—despite slower growth—will still generate 30 % of global trade expansion because of its large existing trade volume. High‑income economies overall are forecast to generate 58 % of trade growth, while low‑ and middle‑income economies contribute 42 %. These figures underline that opportunities exist both in emerging markets and mature economies; diversification should not mean abandoning established partners. AI implication: supply‑chain organisations must adapt to a multipolar world. Geo‑analytics tools can combine demographic data, industrial capacity and policy developments to forecast regional demand and supply patterns. Generative AI can draft bespoke trade reports for executives, summarising the risks and opportunities in each region and recommending engagement strategies. Agent‑based simulations can allocate production across regions to balance cost, risk and sustainability. For instance, an AI‑powered model might suggest splitting manufacturing across India (for cost and scale), Poland (for proximity to European customers) and Kenya (for access to African growth), taking into account transport times, tariffs and carbon footprints.

9. Sector trends: manufactured goods lead, but commodities surge

Manufactured goods remain the largest segment of global trade, but the report highlights that price increases have dramatically boosted the value of certain commodities. From 2017 to 2022, mineral fuels (oil, gas and coal), electrical machinery and equipment, industrial machinery and pharmaceuticals saw the largest increases in trade value. The surge in energy prices following Russia’s invasion of Ukraine amplified the value of mineral fuel trade, while the ongoing digital transformation drives demand for semiconductors and industrial equipment. Pharmaceuticals gained prominence during the COVID‑19 pandemic, exposing the importance of diversified production and resilient cold‑chain logistics. These sectoral dynamics suggest that supply‑chain managers need to monitor not just volumes but also price volatility and regulatory risks. AI implication: generative‑AI models can summarise market intelligence by scanning news, price indices and regulatory updates and generate procurement briefs that highlight supplier disruptions and market movements. Computer‑vision tools can inspect commodities for quality and compliance, verifying the integrity of medicine packaging or detecting surface defects in machinery parts. Reinforcement‑learning agents can negotiate with suppliers by testing different contract terms and adjusting to market conditions. Companies should feed commodity price forecasts into inventory optimisation models to decide when to stockpile critical inputs like chips or pharmaceuticals.

10. Huge headroom for trade growth

Despite decades of globalisation, only 21 % of the value of all goods and services produced worldwide is exported. In other words, nearly four‑fifths of global economic activity remains confined within national borders. Some sectors, such as agriculture and commodities, are highly traded, while others—healthcare, services and construction—remain largely domestic due to regulatory, logistical or cultural barriers. This statistic highlights enormous headroom for future trade growth if policies and technologies lower these barriers. Cross‑border e‑commerce platforms are already enabling small businesses to sell internationally, but obstacles like customs complexity, currency risk and last‑mile logistics still deter many firms. AI implication: there is ample opportunity for AI‑enabled platforms to facilitate cross‑border e‑commerce, match buyers and sellers and automate trade compliance. Recommendation systems can connect exporters with overseas buyers; generative AI can fill out customs forms, harmonise product descriptions and translate documentation; and computer‑vision algorithms can inspect packaging for regulatory compliance. By lowering friction, AI can increase the proportion of goods and services that cross borders. Companies should also monitor emerging digital trade agreements that harmonise data standards and enable smoother flows of both goods and services.

Leveraging AI to Navigate Trade Shifts

The DHL Trade Atlas emphasises resilience and opportunity rather than deglobalisation. To capitalise on these trends, supply‑chain professionals must pair trade intelligence with AI‑driven tools:

• Predictive demand and supply forecasting: Use machine learning to integrate trade‑growth forecasts, macroeconomic indicators and real‑time sales data. For example, incorporate the expected surge in India’s trade volumes into demand forecasts for electronics or pharmaceuticals.
• Multi‑tier risk mapping: Employ graph‑based AI to map suppliers’ suppliers. Identify indirect exposure to Chinese content in U.S. imports and evaluate alternatives.
• Dynamic network design: Use optimisation algorithms and digital twins to simulate different tariff scenarios, route disruptions or regional demand spikes. Compare near‑shoring vs. off‑shoring strategies and assess carbon impacts.
• Generative procurement assistants: Deploy large language models to draft requests for quotation (RFQs), summarise trade‑agreement clauses or analyse political‑risk reports. Generative AI can also localise product documentation for new markets.
• Real‑time geopolitics monitoring: Use natural‑language processing to monitor news and policy updates. Feed alerts into agent‑based systems that adjust sourcing and inventory policies when tariffs or sanctions change.

Hands‑On Exercise: Build a Trade‑Growth Scenario with Generative AI

To make these concepts tangible, try this practical exercise using a generative‑AI tool like ChatGPT or Claude:

Prompt: “Act as an international trade analyst. You have just read the DHL Trade Atlas 2025. Summarise the report’s ten key takeaways in a table with columns: Takeaway, Evidence, and AI‑Driven Action for Supply‑Chain Leaders. For the evidence column, briefly cite statistics or conclusions from the report. For the action column, suggest one way AI could help address or exploit the trend. After the table, propose a five‑year trade‑growth scenario for a company planning to expand sourcing to India and Vietnam, describing how AI‑powered digital twins and multi‑agent systems would manage tariffs, lead times and carbon emissions.”

This exercise pushes the AI to synthesise report insights and translate them into concrete actions. Review the output, adjust the AI‑driven recommendations and share the scenario with your team to spark discussion about future‑proofing your supply chain.

Conclusion

The DHL Trade Atlas 2025 debunks myths about deglobalisation. Global trade is set to continue growing—even under severe tariff scenarios—and opportunities are shifting toward South Asia, Sub‑Saharan Africa and Southeast Asia. Long‑distance trade remains strong, and new growth leaders like India, Vietnam, Indonesia and the Philippines are emerging. To harness these trends, supply‑chain leaders must pair economic intelligence with advanced AI. Predictive analytics can translate macro forecasts into granular demand plans; digital twins and agentic AI can explore alternate futures; and generative models can streamline procurement and compliance. By integrating these technologies, organisations can build resilient, sustainable and competitive supply networks in an era of uncertainty.

References

• DHL Trade Atlas 2025 – The report forecasts that global goods trade will grow modestly faster from 2024‑2029 than in the previous decade and notes that even under proposed U.S. tariff increases, trade remains positive. It warns that U.S. reliance on Chinese content persists despite reduced direct imports. The “Ten Key Takeaways” section highlights that geopolitical shifts are limited and that a complete split of world trade between rival blocs could cut global GDP by up to 7 %; past growth leaders include the UAE, Vietnam and Ireland; future leaders will include India, Vietnam, Indonesia and the Philippines; and standout regions are South Asia, Sub‑Saharan Africa and Southeast Asia. It emphasises that long‑distance trade reached a record 5,000 km in early 2024 and intra‑regional trade fell to 51 %. Only 21 % of global production crosses borders, leaving large headroom for growth.