The AI boom won’t help you scale. Your unit economics will

In Jakarta, a fintech founder used the term "AI-powered" multiple times in a pitch but could not explain how the model made predictions. This highlights a gap between businesses that incorporate AI into decision-making processes and those that merely label their existing operations as "AI-powered." This gap is widening across markets like Colombo, Ho Chi Minh City, and others.

The consensus narrative of an AI boom in Southeast Asia focuses on funding, talent, and AI clusters, but the real story is the concentration of capital in businesses that can demonstrate AI's structural importance, not just its decorative presence. Venture funding has not returned to its 2021 peak, and capital is now consolidating into fewer businesses with AI at their core.

At the micro level, this manifests as a single founder losing deals to a competitor with automated credit scoring. At the SME level, it involves logistics operators falling behind because competitors use AI-driven route optimisation. At the regional level, it includes fintechs clearing licensing faster due to AI-driven compliance.

At the national level, governments are beginning to incorporate AI governance into procurement, favoring vendors with model explainability. This concentration of AI capability into a few businesses creates a moat. Regulatory fragmentation in Southeast Asia is becoming a competitive advantage for businesses that can handle it. Singapore's Model AI Governance Framework is more mature than other countries' approaches.

Founders initially view this fragmentation as a cost center, but those using a scaling framework learn to price it as a capability. For businesses in sectors AI is reshaping, such as fintech, agri-tech, logistics, and healthcare, the path is architectural. They must rebuild their decision layer around the model, not around marketing language.

This requires capital discipline before feature expansion, real governance maturity before a regional launch, and partnerships with regional players who have already solved compliance. For sectors AI has not directly touched, the path is counter-cyclical. Now is the time to acquire AI capability while it is still a specialist skill.

Being early in a hard market with slower AI adoption is an advantage, not a weakness. The key is treating trust and fair terms with customers and partners as a scaling asset, not a compliance checkbox. The businesses that will succeed in five years are those that built governance before regulators demanded it. The pattern shows that AI-native infrastructure plays built around regional constraints are compounding, while AI-flavored consumer plays are not.

Southeast Asia is undergoing an AI-forced audit of which businesses truly understand their unit economics before labeling themselves as "AI-powered." The founders who will scale are the ones who could build without the term and choose to make their business faster, cheaper, and harder to replicate. This is the only AI story worth funding.