Crypto Designed for AI Agents, Not Humans, According to Alchemy CEO

The current financial system was never intended for machine use, having been built around human limitations such as geographical location, sleep patterns, paperwork, and physical presence. However, as AI agents begin to participate in the economy, this human-centric design is becoming a hindrance, according to Nikil Viswanathan, co-founder of the crypto firm Alchemy. "One could argue that cryptocurrency was designed for AI agents, not humans," Viswanathan said. The disparity is evident everywhere: banks have limited hours due to human operation, payments are tied to countries because people reside in them, and credit cards require physical identification and presence. In contrast, AI agents operate without sleep, physical location, or the need to visit banks or use cards, and they are increasingly involved in transactions. "All transactions for agents are online and inherently global," Viswanathan explained in an interview with CoinDesk. This is where cryptocurrency starts to resemble the native infrastructure for a new type of economic actor rather than an alternative financial system. Alchemy, a crypto infrastructure company, provides developers with the necessary tools and services to build blockchain-based applications, including APIs, node infrastructure, and data services that support various products such as financial apps, non-fungible tokens (NFTs), and games. Traditional finance is based on the assumption of friction, where cross-border payments involve currency exchange, intermediaries, delays, and fees, which is normal for humans but unusable for AI agents. Agents require seamless, borderless transactions at any time, often in small increments, and need programmability, direct control over funds via code, and systems that do not rely on physical infrastructure or identity. Cryptocurrency offers exactly that: a global, always-on financial layer where value can be transferred as easily as data. "Cryptocurrency is the global infrastructure for money that agents need," Viswanathan stated. The complexity of cryptocurrency, including seed phrases, private keys, and direct interaction with code, which has long been a challenge for humans, is precisely what makes it powerful for machines. Unlike humans, agents operate natively in code, with "zeros and ones" being their native language, which is also the language of cryptocurrency. For years, cryptocurrency has attempted to become more human-friendly, but its underlying architecture was never designed for humans. Viswanathan compared the shift from crypto tools being built primarily for humans to crypto tools being used by AI agents to the transition from the postal system to the internet. Just as email is more powerful than the postal system because it is designed for computers, cryptocurrency is similar. Looking ahead, AI agents will utilize crypto infrastructure, handling complexity automatically, managing wallets, executing transactions, and optimizing capital flows in real-time, allowing people to control their funds more easily. "You can write code to manage a crypto wallet, but you can't do the same with a bank account," Viswanathan said. The result will be a financial system that is more global, programmable, and autonomous. Viswanathan envisions a layered future, with traditional finance and cryptocurrency as the base, an agent layer operating on top, and a human interface above that. "Just as computers operate the internet and humans use it, agents will operate finance," he said.