In Jakarta, a small seller uses an AI assistant to prepare product descriptions. A student in Manila asks a chatbot to explain a difficult lesson, while a farmer in Vietnam receives data about crops and weather. Across Southeast Asia, artificial intelligence is becoming an everyday tool rather than a distant promise.
Context: AI Arrives Through the Smartphone
Southeast Asia’s AI expansion is being driven by a young, mobile-first population accustomed to adopting new digital services quickly. Research cited by Google in 2025 found that 70 percent of people in the region were already using generative AI tools every week. The technology could contribute an estimated US$270 billion to the regional economy if businesses and governments turn this enthusiasm into productive applications.
Usage is particularly strong in Indonesia. More than 80 percent of Indonesian consumers reportedly use AI tools daily, while approximately 17 million small businesses are adopting the technology to improve efficiency and innovation. For many users, AI appears inside familiar services—search engines, e-commerce platforms, messaging applications and digital wallets—rather than as a separate technical system.
Culture: Teaching Machines to Speak Southeast Asia
The region’s linguistic diversity creates both a challenge and an opportunity. Global models may understand formal Indonesian or English while struggling with Javanese expressions, Thai scripts, Filipino humour, Vietnamese cultural references or conversations that move between several languages.
AI Singapore’s SEA-LION initiative is addressing this imbalance through open-source language models developed around Southeast Asian languages, cultures and contexts. By 2026, the model family had been trained on more than one trillion Southeast Asian-language tokens, attracted over 760,000 downloads and received approximately 180,000 monthly application-programming requests.
Its development reflects a broader regional movement. Vietnam has produced PhoGPT, Thailand is advancing ThaiLLM, and Indonesian technology companies are adapting regional models to understand local vocabulary and consumer behaviour. Better linguistic representation could improve educational tools, public information and customer service while helping smaller communities participate in the AI economy.
Commerce: From Super-Apps to Street Stalls
AI is already embedded in the services used to move people, products and money. Grab and Gojek apply predictive systems to match passengers with drivers, plan delivery routes, identify suspicious transactions and estimate demand across crowded cities.
For microbusinesses, generative AI can prepare advertisements, translate customer messages, organize orders and create product photographs. A home-based food producer in Surabaya can write promotional material without hiring an agency, while a guesthouse in Chiang Mai can communicate with international visitors more efficiently.
In the Philippines, where business-process outsourcing remains a major employer, companies are introducing AI for transcription, quality control and customer assistance. The challenge is to help workers move from repetitive tasks toward supervision, problem-solving and higher-value services.
Regional applications are also expanding beyond commerce. Google reported that more than 85,000 Southeast Asian researchers had used the AlphaFold database for biological research. Agricultural AI tools are being introduced in Indonesia, Malaysia and Vietnam to analyse crop types, field sizes and planting cycles.
Concern: The Gap Behind the Enthusiasm
High adoption does not guarantee equal benefits. The ASEAN Digital Outlook 2026 found that Singapore, Malaysia, Thailand and Brunei performed strongly in AI-related data and infrastructure readiness. Indonesia, the Philippines and Vietnam remained slightly below the regional benchmark, while Cambodia, Laos and Myanmar faced wider gaps.
Those differences can determine who receives accurate information, develops valuable skills or finds employment in an AI-supported economy. Rural communities may lack affordable connectivity, while models trained on limited data can misrepresent minority languages and local conditions.
Privacy, copyright, biased automated decisions and AI-generated fraud are equally urgent concerns. Deepfakes can imitate public officials, celebrities or family members with increasing credibility. Filipino AI specialist Dr Erika Fille Legara warned, “Without governance, AI exposes people to very real harm—discrimination, misinformation, exclusion. Trust can collapse.”
Choice: Building AI People Can Trust
ASEAN has begun creating a shared foundation through its Guide on AI Governance and Ethics, the Expanded Guide on Generative AI and the Responsible AI Roadmap. These frameworks promote accountability, security, content verification and human oversight while allowing countries to develop policies suited to their circumstances.
Governments must now translate principles into protections that people can understand and use. Schools need practical AI literacy, workers require accessible retraining, and businesses should disclose when automated systems influence important decisions. Developers must involve local speakers and communities when building regional datasets.
ASEAN Secretary-General Dr Kao Kim Hourn expressed the ambition clearly: “This is our chance to co-create a future where AI serves humanity with purpose.”
Southeast Asia’s greatest advantage is not simply its willingness to adopt AI. It is the region’s capacity to shape technology around its languages, communities and everyday realities. The next chapter will depend on ensuring that innovation expands opportunity without leaving trust, dignity or vulnerable users behind.
