Cathie Wood advises smart investors to monitor AI agents’ spending habits

As AI agents shift from answering queries to spending actual funds, investors and tech giants are racing to dominate the financial networks that fuel machine-driven commerce.

— AI agents are evolving from answering questions to taking actions and spending money, prompting new questions about the financial infrastructure they will rely on.

— Stablecoins and blockchains could serve as payment rails for agents purchasing data, computing power, and other digital services, experts suggest.

— The key debate now is whether agentic finance will develop on open networks or become concentrated among a few banks, payment providers, and technology platforms.

Cathie Wood has long urged investors to “follow the developers” to predict technology trends. Now, the ARK Invest CEO believes investors should instead focus on “the agents.”

“We’re probably going to be talking more and more about ‘follow the agents,’” Wood said during a panel at Robinhood’s Summit in Houston on Wednesday. She was referring to AI agents—software that can perform tasks on behalf of a person, not just answer questions or generate text.

Wood’s brief remark, near the end of a broader discussion on AI, private markets, and tech investing, highlights a growing concern as companies race to build AI agents: What happens when AI stops just providing answers and starts spending people’s money?

Developers have historically indicated which technologies are gaining traction because engineers tend to adopt tools they find useful. If millions of AI agents start making their own decisions about which software, services, and networks to use, their activity could reveal new demand trends.

However, these agents will also need a way to pay.

In a recent post, Joseph Chalom, co-CEO of SharpLink and former head of digital assets at BlackRock, warned that the financial system used by AI agents should not be dominated by a few banks or tech companies.

“A world full of intelligent agents means nothing if a handful of companies decide where your money can go,” Chalom wrote in the final part of a three-part series on agentic finance.

Chalom believes the issue goes beyond whether an AI agent can spend money. It’s about how much control people give the agent and who oversees the financial system behind it. For example, a user might authorize an agent to spend up to $500 on hotel bookings without granting unlimited access to their bank account. Users should also be able to revoke that authority and view a record of the agent’s actions.

Chalom also argued that users should be able to transfer their agents between financial providers instead of being locked into a single company’s system. An agent should be able to carry its identity, financial information, and permissions to another provider, similar to how a person can switch phone carriers while keeping their number.

He sees a role for crypto in this context.

Chalom mentioned open blockchains like Ethereum.

This also adds a financial dimension to Wood’s idea of “following the agents.” If agents start handling more tasks independently, investors could monitor not only which AI models and software they use but also how they make payments and which financial networks they rely on.

BlackRock highlighted a similar link in a September paper exploring the intersection of AI and digital assets. The firm argued that AI agents could drive new demand for machine-oriented payment systems. An agent might need to pay for an API call, purchase data from another service, or rent computing power—all without waiting for human approval of each transaction.

Stablecoins and blockchains could serve as one method for handling these payments, BlackRock noted.

Stablecoins operate 24/7, and blockchain-based payment protocols enable software to send small payments directly to other software. For instance, Coinbase’s (COIN) x402 is designed to allow machines to pay for online services like data or API access.

There are already signs of agents emerging in other crypto areas. Coinbase CEO Brian Armstrong posted on X that “Grok is the leading client for agentic traders on Coinbase currently,” though he didn’t provide specific figures or additional details about the activity.

Crypto won’t have this market to itself. Stripe, Visa, Google, and OpenAI are among the companies developing ways for agents to make purchases, and BlackRock emphasized that traditional payment systems will remain important.

This competition could make Wood’s advice to “follow the agents” particularly valuable for crypto investors. If AI agents become major economic players, monitoring their transactions could provide a new way to assess whether stablecoins and blockchains are gaining real-world adoption or if most agent activity remains on traditional payment networks.

As stablecoins enter regulated finance, the APAC region is emerging as a critical testing ground. This report outlines the region’s regulations, use cases, and the role of RLUSD.

Why it matters:

As stablecoins enter regulated finance, the APAC region is emerging as a critical testing ground. This report outlines the region’s regulations, use cases, and the role of RLUSD.