Turning displays of artificial intelligence into everyday productivity

The 23rd China-ASEAN Expo in Nanning, Guangxi Zhuang autonomous region, features an artificial intelligence pavilion showcasing applications in various sectors, including manufacturing, smart cities, medical care, finance, cybersecurity, agriculture, education and business matching.

The key question is what happens after visitors leave the exhibition hall. Can an AI matchmaking system help a small exporter find reliable partners across borders? Can affordable AI tools help a family business handle customer messages, inventory and digital payments?

A realistic answer begins with a simple truth: AI creates value only when it is localized, trusted, and embedded in everyday work. This is especially true in Southeast Asia, where economic integration depends not only on infrastructure and market access but also on language diversity, small-business capacity and digital safety.

Language is the first bottleneck. The Association of Southeast Asian Nations is a 670-million strong market, characterized by a young population, rising internet penetration and rapidly growing digital services. It is also one of the world's most linguistically diverse regions, with more than 1,200 languages. For AI, this diversity is not merely cultural. It forms the technical and social environment in which systems must operate.

This is why AI translation devices need careful consideration. While they reduce the friction of first contact, which is crucial in trade, tourism and education, cross-border commerce requires accuracy in contracts, product documentation, compliance requirements, warranties and dispute resolution. Improving the translation of low-resource languages requires more data, adapting models to specific domains and enhancing quality through bilingual examples.

This doesn't imply translation devices aren't useful, but that they are merely the visible endpoint of a deeper system. Southeast Asia needs a shared language infrastructure: multilingual trade corpora, tourism-service datasets, education content, agriculture advisory terms, cybersecurity lexicons and evaluation benchmarks. Without these resources, AI tools may work well in formal or high-resource settings while failing in practical contexts where they are most needed.

Recent efforts in Southeast Asian language models indicate that this challenge is solvable, though not automatically. The broader lesson is that AI becomes useful when trained and adapted to the languages, scripts, cultural norms and practical situations of its users.

The second issue is the adoption of AI by micro, small and medium-sized enterprises. An OECD report on ASEAN argues that AI can raise productivity, improve supply-chain management and lower trade costs, but warns that ASEAN's AI readiness is uneven, constrained by data governance, access to AI goods, services restrictions and specialist mobility.

This matters because micro, small and medium-sized enterprises are not marginal; they account for 97 percent of ASEAN businesses and 85 percent of employment.

Inclusive AI doesn't mean simply making advanced tools available online. Adoption depends on access, skills, awareness and trust.

AI can support customer engagement, operations, analytics and decision-making, but smaller firms in developing countries face persistent barriers in finance, skills, technology, privacy and cybersecurity. In practical terms, adoption requires low-cost tools, sector-specific templates, local-language interfaces, training through business associations and platforms, and clear safeguards for data and liability.

The third issue is safety. Digital connectivity aids e-commerce, but also enables cross-border harm.

In 2025, the UN Office on Drugs and Crime warned that cyber fraud networks in the Mekong region had become an interconnected ecosystem, with hundreds of industrial-scale scam centers generating almost $40 billion in annual profits. Criminal syndicates were integrating malware, generative AI and deepfakes into cyber-enabled fraud, causing substantial financial losses across East and Southeast Asia.

Here, too, the answer lies in institutional cooperation rather than technological optimism. The 2026 ASEAN Guide on Anti-Scam Policies and Best Practices argues that no single country can tackle scam calls and SMS alone, because scammers exploit cross-border coordination gaps and national safeguard differences. It recommends harmonized baseline measures, stronger regional collaboration and coordinated responses to scam calls and messages.

The foundation for this already exists. In 2023, ASEAN, China and UNODC agreed to jointly address transnational organized crime and trafficking.

The next step is to make anti-fraud AI a cooperative public good: shared scam typologies across languages, privacy-preserving exchange of threat indicators, joint testing of deepfake-detection tools, faster channels among platforms, telecom firms, banks and law enforcement, and public education in local languages.

The China-ASEAN Countries AI Application Cooperation Center, launched in Nanning in 2025, could play a useful role by converting demonstrations into pilots, testbeds and capacity-building programs. But the criterion for success should be modest and practical.

The question is not how impressive AI looks in a pavilion, but whether it lowers the ordinary costs of translation, coordination, compliance, customer service and fraud prevention for people and firms across Southeast Asia.

Seen this way, the future of China-ASEAN digital connectivity is not only in data centers, trade agreements or exhibition halls.

It lies in everyday languages, small-business routines and trusted safeguards.

AI will matter most when it becomes less of a spectacle and more of a reliable layer in the work of trading, learning, traveling and protecting one another.

Duy Dang-Pham is the associate head of Department (Research and Innovation), at The Business School of RMIT University Vietnam; and Thanh-Thuy Nguyen is a senior lecturer at the School of Accounting, Information Systems and Supply Chain, College of Business and Law, RMIT University, Australia.

The views don't necessarily reflect those of China Daily.

If you have a specific expertise, or would like to share your thought about our stories, then send us your writings at opinion@chinadaily.com.cn, and comment@chinadaily.com.cn.