This change has already begun. Robinhood launched AI-powered investing tools in May that allow agents to trade stocks and make purchases for users. CEO Vlad Tenev said AI agents will eventually rival the capabilities of human traders, while OpenAI and Anthropic strive to build increasingly autonomous systems that can navigate software and complete complex tasks on their own.
For Kaul, these agents pose a problem that current payment systems are not designed to solve.
Many transactions between AI agents might be worth only fractions of a cent, such as paying for an API call, a second of computing power, or access to a dataset. Traditional payment networks become expensive when the fees cost more than the transaction itself.
This is where Kaul thinks blockchains come into play.
She argued that public blockchain networks are better suited for machine-to-machine payments because they offer programmable transactions, cryptographic identity and near-instant settlement. Instead of relying on banks or card networks, AI agents could hold digital assets and pay themselves directly through blockchain rails.
If this happens at scale, demand for blockchain networks could grow alongside AI adoption.
Since agents would need native cryptocurrencies to pay network fees, Kaul argued that increasing transaction volumes could increase demand for these tokens while generating more revenue for developer incentives, network security, and decentralized applications.
