Coinbase CEO Brian Armstrong has reiterated his strong belief that crypto and stablecoins will serve as the primary payment rail for artificial intelligence agents.
In a post on X, Armstrong responded to a note about BlackRock highlighting AI agents as a potential major source of stablecoin demand.
He agreed by outlining three simple points: the number of AI agents will continue to grow, more of those agents will need to transact, and crypto along with stablecoins will become their preferred payment method.
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
This view is not new for Armstrong. Throughout 2026 the Coinbase CEO has consistently argued that artificial intelligence and crypto are complementary rather than competing forces.
The idea that crypto and stablecoins could become the go-to payment method for AI agents is gaining traction as AI systems move from simply answering questions to acting independently on behalf of users and businesses.
AI agents are becoming economic actors. Notably, the next phase of AI is not just about chatbots generating text or answering questions. AI agents are increasingly being designed to take actions autonomously like booking flights, purchasing products, accessing databases, buying computing resources, paying for APIs, and even executing financial transactions.
Visa describes this emerging model as agentic commerce, where AI agents can search, select, and purchase products or services on behalf of people or businesses.
Also, AI provides programmable intelligence and labor, while crypto supplies programmable money and markets. Together, Armstrong says, they create what Coinbase calls “Agentic Finance” or AiFi. Armstrong has repeatedly stated that AI agents will eventually conduct far more transactions each day than all humans combined.
These autonomous systems cannot open traditional bank accounts, wait days for wire transfers, or easily navigate national banking systems and KYC requirements designed for people. They need real-time, global, low-cost, programmable money—and that, in his view, is precisely what crypto and stablecoins deliver.
Coinbase has spent significant resources building the infrastructure to support this future. The company points to its Base blockchain, the USDC stablecoin, and the x402 payment protocol (which revives the long-unused HTTP 402 “Payment Required” status code) as core tools already powering the majority of on-chain agentic payments.
Coinbase has also developed agent-specific wallets, tools that let AI agents connect to user accounts or hold their own balances, and features that allow agents to trade crypto, stocks, and other assets within set guardrails.
In recent months the platform has expanded so agents can book travel, make micropayments, and execute financial tasks autonomously using USDC on Base.
Armstrong’s thesis rests on practical advantages. Many agent-driven transactions are extremely small often under 30 cents, making traditional card networks inefficient because of fixed minimum fees.
Crypto rails can settle these micropayments almost instantly for fractions of a cent. Agents can also hold funds, pay other agents for specialized services, raise capital, manage portfolios, and handle routine financial chores such as bill payment or tax-related tasks without constant human intervention.
The broader implication is that a large new class of economic actors software agents operating at machine speed, could drive substantial demand for stablecoins and on-chain activity.
BlackRock’s observation about potential stablecoin demand from AI agents aligns with this outlook. While traditional finance remains essential for human users, Armstrong positions crypto as the natural financial layer for the emerging agentic economy.
Coinbase continues to position itself at the center of that shift by offering the accounts, payment rails, and trading tools these agents will require.
Looking ahead
The outlook for agentic payments will likely depend on how quickly AI agents move from experimental tools to widely deployed economic actors.
As agents become capable of independently purchasing data, software, computing power and other digital services, the volume of machine-to-machine transactions could rise substantially.
Stablecoins could benefit from this shift because their dollar denomination, programmability and ability to settle transactions on blockchain networks make them suitable for automated payments.
However, their eventual role is likely to depend on factors including transaction costs, regulatory requirements, security, interoperability, and whether traditional payment networks can provide comparable infrastructure for autonomous transactions.