The survey, the first to focus on business adoption of AI for trade, found large gaps in AI adoption between micro, small and medium-sized enterprises and large firms as well as between firms in high-income and low/lower middle-income economies. In high-income economies, 66 ­per ­cent of firms reported having adopted AI, compared to 27 ­per ­cent of firms in low/lower middle-income economies. 

The survey further found that large firms primarily use AI for trade compliance, while micro, small, and medium enterprises (MSMEs) focus on market intelligence and communications. Firms in low and lower-middle-income economies moreover said they leverage AI to grow their international trade. This highlights AI's potential to foster inclusive participation in global trade by improving access to market intelligence, trade compliance tools, and knowledge of preferential trade agreements, helping bridge the AI adoption gap.

Among those firms using AI, the survey found that close to 90 ­per­ cent reported benefits in trade-related activities. Specifically, 22 ­per­ cent saw gains in efficiency and productivity, 14­ per­ cent observed better decision-making, 10 per cent enjoyed an expanded export base and 17 per cent widened their range of traded import and export products. The survey also shows that there is potential for AI to help firms navigate complex trade rules and to benefit more from trade agreements. Three-quarters of responses indicated that firms were using AI for various customs-related applications, and 20­ per­ cent indicated they use AI to pinpoint trade compliance risks.

 The survey also identified businesses' concerns about AI and the barriers to its adoption. Firms reported regulatory uncertainty and fragmentation, particularly of data policies as key challenges.

Nearly one fifth of responses (19 ­per ­cent) cited regulatory uncertainty as a top concern. Meanwhile, 12­ per­ cent highlighted regulatory or compliance challenges as a key barrier. In addition, 7 ­per ­cent cited regulatory heterogeneity as a challenge in adopting AI, which shows how regulatory fragmentation can create obstacles to the cross-border adoption of AI. A third of the responses on AI-related cross-border regulatory challenges pointed to adapting to different data privacy requirements as the most significant issue. Sixty-four per cent of AI-adopting firms, furthermore, said they expect at least 5 per cent impact on their compliance costs due to differing data protection regulations.

Businesses also cited technical readiness as a concern: 14 ­per ­cent of responses reported difficulty accessing high-quality data for training AI systems, while another 14­ per­ cent cited limited AI expertise or skills as a challenge.

The full report is available here. Some of the early findings were reported in the 2025 World Trade Report.