Two economists from the ASEAN+3 Macroeconomic Research Office (AMRO) believe that the winner in the artificial intelligence (AI) race will not be the country that develops the most powerful model, but the one that manages to establish the currency on which these models and their accompanying infrastructure are based.
Chancxu Fu and Xiago Huang recently suggested that the strengthening of an economic cycle, where the energy for AI data centers, AI infrastructure, and the very cost of using AI are valued in US dollars, could lead to an increase in the dominance of the US dollar through stablecoins.
The thesis underlying this premise is simple: if companies have to pay for everything AI-related in dollars, and AI becomes a much larger industry than today in terms of monetary turnover and significance, there will be a renewed sharp increase in global demand for dollar liquidity. The authors define this factor as the first channel of dollar dominance in AI.
The second channel is related to the use of a specific currency that will facilitate payments involving agents; such payments could become ubiquitous if AI, as many in the crypto industry believe, consistently displaces human-managed systems in logistics, inventory management, and treasury operations. "In particular, dollar-pegged stablecoins could provide the programmable settlements necessary for agent-based commerce," they explain.
These two channels can intersect, as agent systems could execute settlements for AI-related computational payments in stablecoins, creating a positive dollar loop, from which it is difficult to break free.
Although stablecoins are not the only available substitute for the dollar, they could benefit from early network effects, as they are already ready to play this role in the AI economy, while tokenized deposits of central bank digital currencies (CBDCs) are not yet ready for this task.
This would have a side effect that would also benefit the US: increased demand for US Treasury bonds used as collateral to support the growing market capitalization of stablecoins.
The report calls on ASEAN+3 countries, including Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines, Singapore, Thailand, Vietnam, China, Japan, and South Korea, to limit their dependence on this "dollar loop."
Creating regional data centers and developing tokenized forms of money based on local currencies could help avoid strengthening the link between stablecoins and AI, as well as help these countries participate in the AI economy without indirectly supporting a new level of dependence on AI.
