Markets Cathie Wood suggests that savvy investors should pay close attention to the spending behaviors of AI agents.

The shift of AI agents from simply responding to inquiries to actively spending money is prompting investors and tech companies to focus on the financial systems that will facilitate this machine-driven commerce.

By Helene Braun, AI Boost|Edited by Cheyenne Ligon Oct 3, 2026, 11:00 a.m. EDT 4 min read

Key Points:

  • AI agents are transitioning from merely answering questions to making purchases, raising concerns about the financial systems they will utilize.
  • Experts believe stablecoins and blockchains could serve as the payment infrastructure for agents acquiring data, computing resources, and other digital services.
  • The emerging competition will determine whether agent-driven finance operates on open networks or becomes dominated by a few banks, payment processors, and tech firms.

Cathie Wood, CEO of ARK Invest, has long advised investors to “follow the developers” to gauge technological trends. Now, she believes it may be time to shift that focus to the agents themselves.

“We’re probably going to be talking more and more about ‘follow the agents,’” Wood stated during a panel discussion at Robinhood’s Summit in Houston. She was highlighting AI agents, which are software programs capable of performing tasks on behalf of individuals rather than just providing responses or generating text.

While her comment was brief and came towards the conclusion of a broader discussion about AI, private markets, and tech investments, it underscores a critical question: What will occur when AI agents transition from simply delivering answers to actually spending money?

For a long time, developers have been instrumental in indicating which technologies are gaining traction, as engineers typically favor tools they find beneficial. If millions of AI agents begin making independent decisions about the software, services, and networks they utilize, their behaviors could provide insights into emerging demand.

Yet, these agents will require a mechanism for payment.

In a recent post, Joseph Chalom, co-CEO of SharpLink and former digital assets chief at BlackRock, emphasized that the financial infrastructure for AI agents should not be monopolized by a limited number of banks or tech firms. “A world full of intelligent agents means nothing if a handful of companies decide where your money can go,” he remarked in the final part of a series on agentic finance.

Chalom argues that the key issue is not just whether an AI agent can spend money, but also the extent of control individuals grant these agents and who manages the underlying financial infrastructure. For instance, a user might permit an agent to spend up to $500 on a hotel booking without giving it unrestricted access to their bank account. Users should also retain the ability to revoke such permissions and track the agent's transactions.

Moreover, Chalom believes individuals should have the freedom to transfer their agents between financial institutions rather than being confined to a single provider’s ecosystem. An agent should be able to carry its identity, financial details, and permissions across to another provider, similar to how a person can switch phone carriers while retaining their number.

Chalom envisions a role for cryptocurrency in this scenario.

He suggests that open blockchains like Ethereum ETH$2,681.93 could offer a shared financial network accessible to various agents, applications, and companies. Instead of each AI firm developing its own proprietary payment system, agents could move funds across a collective network without necessitating a single bank or tech company to mediate every transaction.

This perspective enhances Wood's notion of “following the agents” with a financial aspect. As agents undertake more autonomous tasks, investors could observe not just which AI models and software they select but also how they finance their activities and which financial networks they opt for.

BlackRock has drawn similar conclusions in a September report exploring the intersection of AI and digital assets. The asset management firm posited that AI agents could drive demand for payment systems designed for machine use. An agent might need to pay for an API call, purchase data from another service, or rent computing power without requiring human approval for each transaction.

According to BlackRock, stablecoins and blockchain technology could facilitate these payments.

Stablecoins can operate continuously, while blockchain-based payment protocols allow software to make minor payments directly to other software. For example, Coinbase's (COIN) x402 is designed to enable machines to pay for online services like data or API access.

There are indications of agents emerging in other sectors of cryptocurrency as well. Coinbase CEO Brian Armstrong noted on X that “Grok is the leading client for agentic traders on Coinbase currently,” although he did not disclose specific figures or further details on the activity.

However, the cryptocurrency space will not hold this market exclusively. Companies like Stripe, Visa, Google, and OpenAI are also working on solutions for agents to make purchases, and BlackRock has asserted that traditional payment systems will remain significant.

This competitive landscape makes Wood's advice to “follow the agents” particularly pertinent for crypto investors. If AI agents become significant economic players, tracking their transactions could provide a novel method to assess whether stablecoins and blockchains are achieving practical utility or if most agent activities remain tied to conventional payment systems.

Artificial Intelligence AI Disclaimer: Parts of this article were generated with the assistance of AI tools and reviewed by our editorial team to ensure accuracy and adherence to our standards. For additional information, see CoinDesk's full AI Policy.