As AI agents begin carrying out transactions as economic actors, a claim has emerged that crypto infrastructure is more suitable than traditional finance.
On April 25, blockchain media outlet CoinDesk reported that Alchemy co-founder and CEO Nikhil Viswanathan (니킬 비스와나탄) said crypto has a structure better suited to AI agents than humans.
He said the modern financial system is designed around human living conditions. Bank business hours follow human life cycles, payments are divided by country, and credit cards assume physical identity and face-to-face interaction. AI agents, by contrast, do not sleep, do not belong to a specific country, and do not visit banks or carry cards.
Viswanathan said the difference becomes clearer in transactions. AI agents transact entirely online and are inherently global, but traditional finance assumes currency exchange, intermediaries, delays and fees, making it hard for agents to use, he said.
Alchemy is a crypto infrastructure company that provides APIs, node infrastructure and data services needed to develop blockchain-based applications. It helps companies building on-chain products, from financial apps to non-fungible tokens and games, avoid handling the complexity of operating blockchains directly.
He said AI agents must be able to transact in small amounts across borders at any time and must be able to directly control funds through code. He described crypto as a global money infrastructure that meets those conditions.
He also stressed that characteristics of crypto that were inconvenient for humans can be strengths for machines. Seed phrases, private keys and direct interaction with code are barriers to entry for ordinary users, but are natural for AI agents operating in a code environment, he said.
Viswanathan said crypto has been refined to appear human-friendly, but its underlying structure was not designed around humans from the start. He compared it to email, saying it is more powerful than mail because it was designed for computers, and said crypto has the same nature.
He laid out a structure in which traditional finance and crypto form the base layer, and AI agents on top automatically handle wallet management, transaction execution and optimization of cash flows. Humans would control funds more easily through the interface above, and it could ultimately develop into a way in which agents run finance, he added.