Google-led report forecasts $368 billion Southeast Asian digital economy as AI use grows

Southeast Asia’s digital economy is projected to reach $368 billion in transaction value in 2026, up 17% from last year, according to a new Google, Temasek and Bain & Company report. Its next growth proposition is more ambitious: moving consumers from asking AI for advice to letting it complete purchases on their behalf.

Published October 21, the 11th annual e-Conomy SEA report, titled Building the Intelligent Economy, puts widespread AI adoption at the center of its outlook. It estimates that 243 million people actively use generative AI, equivalent to 58% of the region’s working population. Adoption is scaling four times faster than mobile internet did, the report says.

The $368 billion forecast measures gross merchandise value, or transaction value, rather than the revenue earned by digital businesses. The authors credit widespread AI implementation with helping sustain that growth and project that the region’s digital economy will exceed $1 trillion by 2035.

Shopping assistants lead the commercial shift

E-commerce anchors the forecast, with projected transaction value of $226 billion this year, up 18%. The report describes shoppers using search across different media and in-app assistants to navigate complex purchases. It associates those personalized experiences with 10% higher purchase conversion rates and increases in order volume as large as 75%. Those are report-supplied performance figures, not guaranteed results for every retailer.

Travel shows how AI is changing discovery before taking over transactions. Traditional search engines remain the leading planning tool, but generative AI has overtaken social media as the second-most-used option for organizing complex itineraries. The report forecasts $59 billion in online travel transaction value in 2026, an increase of 10%.

Elsewhere, the authors point to predictive suggestions and smarter routing in mobility and food delivery. In digital financial services, they identify real-time fraud detection, broader access to credit and personalized recommendations that complement human advice. These are different uses of AI, spanning recommendations, operational decisions and consumer protection.

Permission to buy comes with conditions

The proposed next step is agentic commerce: AI systems comparing prices and completing transactions for users. The report says 72% of users already rely on automation such as price tracking, while 75% express willingness to delegate comparisons and purchases to AI agents. Willingness is not the same as completed autonomous purchases.

That openness sits alongside rising confidence: 62% report greater trust in AI tools than a year earlier. Proven track records, familiarity and output accuracy are the leading reasons cited. But consumers also identify specific protections they want before becoming comfortable with new agent capabilities, including a way to undo transactions.

44%Privacy safeguards

Users identify robust privacy safeguards as a condition for comfort with new agent capabilities.

36%Clear capabilities and limits

Users want transparent information about what AI can and cannot do.

30%Reversible transactions

Users identify the ability to reverse agentic transactions as a safeguard they expect.

The next phase needs more than consumer trust

The report projects a 60-fold increase in regional AI usage by 2030. Supporting that demand would require roughly three times the current data-center capacity over the next five years, with AI workloads taking a dominant share of infrastructure. Its infrastructure requirements extend beyond building more facilities to how efficiently they operate and how they obtain energy.

• Affordable operation: Running AI models must keep getting cheaper and deliver practical value so businesses of different sizes can sustain adoption.
• Useful autonomy: Businesses need to move beyond single-question answers toward completing multi-step tasks, backed by privacy protections and clear safety rules.
• Energy discipline: The authors call for stronger model and facility efficiency, alongside expanded access to clean energy, to support the computing buildout.

Regional rules are another part of the proposed foundation. The report points to the ASEAN Digital Economy Framework Agreement, expected in November 2026, as a planned region-wide digital trade pact. It presents unified governance as one of the conditions needed to support the projected expansion, alongside infrastructure, efficiency and sustainable AI economics.