For decades, the “wealth gap” in emerging markets has been, at its core, an “expertise gap”. In countries like the Philippines, Indonesia, and Thailand, the barrier to wealth creation isn’t just a lack of capital but also a lack of access to the sophisticated financial services and personalized advice that are available to the wealthy.
In the mid-2010s, a promise echoed across Southeast Asia: Bitcoin and decentralized finance would be the “great equalizers”. The belief being that by cutting out the gatekeepers of traditional finance, wealth would start to flow to the unbanked and the under-leveraged.
The reality has been more nuanced. In Vietnam, where over 20% of the population owns digital assets, crypto has been a double-edged sword. While it has expanded access to alternative financial rails, it has also introduced new forms of wealth concentration. Wealth in decentralized networks remains highly concentrated. Some studies suggest that as little as 0.01% of holders account for over a quarter of Bitcoin in circulation. The promise of equal access has not yet translated into equal outcomes.
As we now experiment with AI financial agents, we stand at a far more dangerous crossroads. If we do not address the structural barriers of access and education, we might end up automating inequality even more.
How can AI agents narrow the wealth gap?
If AI agents are to close the wealth gap, they must first remove the barrier that is paid expertise. In many emerging markets, the rich pay for tax advisors, asset managers, and may even have family offices to optimize their investments. On the other end of the spectrum, we have segments of the population who are penalized for the lack of it, losing an estimated 5% - 10% of their annual income to hidden fees and exorbitant spreads.
In theory, AI agents could narrow the wealth gap by reducing the cost of premium financial advice to the price of a data connection and a subscription. For instance, an AI agent for a Filipino freelancer could automatically sweep tiny amounts of spare change into the highest-yielding, risk-adjusted DeFi pools. Or in the case of remittances, an AI agent could be set to monitor which specific stablecoin pair is more favorable, execute a swap and route the funds through the most cost-effective rail.
While these scenarios sound like great entry points for AI agents, we need to ask another question: can these consumers get into the room in the first place?
Do the underserved have a seat?
The reality is, AI is not currently a universal utility. For an AI agent to function, consistent, high-speed data and a device capable of running sophisticated interfaces are required. In many rural parts of Southeast Asia, data is still a metered luxury, and so is a smartphone advanced enough to navigate complex DeFi protocols. Also, the best performing agents that have access to sophisticated risk models are the same ones locked behind $20/month subscription fees. These are out of reach for a street vendor in Da Nang or a taxi driver in Bali.
Further, in the world of agentic finance, latency (the “waiting time” between a command and a result) could see rural players picking up scraps left behind by “alpha agents” owned by institutional investors from financial hubs in the cities. All this means the rich, once again, find a way to become richer and the poor, no better off.
But perhaps the most overlooked factor is that AI agents must be paired with access to higher learning or specialized digital literacy. Using an AI agent efficiently requires a level of financial knowledge and critical thinking that education systems are not providing. The belief that AI “removes the need to think” cannot be more wrong. If a user cannot critically evaluate the solutions proposed by AI, they are setting themselves up to be sabotaged by actors looking to exploit such loopholes.
The Next Great Divergence
If efficient and cost-effective AI agents remain out of reach, we face what the UNDP (United Nations Development Program) calls the “ Next Great Divergence ”. If we build AI agents without addressing the structural cracks, we are simply automating the status quo. Your gift of a Ferrari to a family is useless unless the roads around their house are paved.
As DeFi builders, our job is to pave these roads, building the decentralized, transparent, and low-cost infrastructure that allows AI to serve the many, not just the few who can afford a subscription. An open-source and decentralized interface that enables AI agents to interact with payments onchain will allow a builder in the Philippines to build a hyper-local financial tool for their community without being “taxed” or shut down by a centralized gatekeeper.
The first wave of crypto expanded access to alternative assets but not always the tools needed to use them effectively. AI agents could be our best shot at closing the wealth gap — but only if we build infrastructure that gives everyone a stake. A warung owner's digital CFO should have the same analytical prowess, speed, and strategic depth as the tools used by global fund managers.
Colin Goltra is the chief executive officer of Morph , a blockchain platform building universal infrastructure for boardless payments and financial services. A veteran of the global crypto industry, Colin has been an early adopter and advocate for digital assets throughout his career. Before joining Morph, he served as chief operating officer at Yield Guild Games (YGG) and as director of Southeast Asia at Binance, where he played a key role in driving regional growth and ecosystem development. Colin has been in the global crypto ecosystem for 12 years and remains an avid believer in technology and its ability to improve the world. Prior to crypto, his background was in traditional technology and finance in Silicon Valley.
