Coinbase CEO Pushes Back on AI Pivot, Says Crypto Will Power AI Agents in New Financial Era

Coinbase CEO Rejects Zero-Sum Thinking on Crypto and AI

Brian Armstrong, Coinbase CEO, has pushed back forcefully against the notion that crypto firms pivoting to AI represents the right strategic move. In a post on X Sunday, Armstrong called the "pivot to AI" advice a case of "zero-sum thinking" that fundamentally misunderstands how these technologies interact 1

1

. "Crypto is a general purpose technology. It's infrastructure, the same way electricity or the internet is infrastructure," Armstrong wrote, arguing that it doesn't compete with the next big thing because it underpins it 3

3

. Rather than viewing the relationship as either-or, Armstrong positioned crypto and AI as complementary technologies where AI agents driving crypto adoption will create unprecedented demand for blockchain-based financial services.

AI Agents Need Programmable Money for Financial Infrastructure

The core of Armstrong's argument centers on AI agents and their unique requirements for financial infrastructure. "AI agents will need their own financial infrastructure and will eventually transact far more per day than all humans combined," Armstrong stated 2

2

. Unlike humans, AI agents cannot open bank accounts, wait three days for wire transfers, or operate within a single country's banking system. They require real-time programmable money to hold funds and execute payments autonomously 1

1

. Armstrong envisions these agents engaging in trading, acting as financial advisors, raising or borrowing money for projects, and eliminating human tasks around tax planning, portfolio rebalancing, and bill payments 4

4

. This vision, which Armstrong branded "Agentic Finance" or AiFi, positions crypto as the essential money rails while AI supplies the intelligence layer.

Base Network and x402 Protocol Drive Automated Machine-to-Machine Payments

Coinbase has built significant infrastructure to support this symbiotic relationship between crypto and AI. The company's Base network, launched in 2023 as an Ethereum layer-2 network, was designed to make onchain applications faster and cheaper to use 2

2

. Agentic payment activity on Base topped 100 million transactions in June, according to Chainalysis, which tracked x402-related payment flows onchain 2

2

. The x402 protocol, introduced by Coinbase, revives the long-unused HTTP 402 "Payment Required" status code and transforms it into a payment protocol enabling automated stablecoin payments between software applications 4

4

. The USDC stablecoin, launched by Circle and the Coinbase-backed Centre Consortium in 2018, powers "the vast majority" of agentic payments according to Armstrong 1

1

.

Coinbase Becomes AI-Native While Building Crypto Rails

Armstrong's vision arrives after Coinbase itself leaned heavily into AI transformation. Earlier this year, the exchange cut 14% of staff amid a crypto downturn, in a restructuring Armstrong framed as a drive to rebuild the company "lean, fast, and AI-native" with "AI at our core" 1

1

. Rather than choosing between technologies, Armstrong positioned AiFi as Coinbase doing both simultaneously—using AI internally while constructing the crypto infrastructure that agents will transact on. Calling the moment an "inflection point" in an email to staff, he argued the company was "adjusting early and deliberately to rebuild Coinbase" 1

1

. The timing is notable as Coinbase prepares to report second quarter earnings Thursday, with analysts expecting revenue of $1.29 billion and a 13.8% decline compared to last year's period 2

2

.

Growing Momentum Across AI Agent Ecosystem

The broader AI agent ecosystem shows accelerating adoption that supports Armstrong's thesis. OpenAI and Anthropic have both highlighted growing use of autonomous AI agents, with Anthropic reporting increased adoption of agentic AI tools like Claude Code and Cowork for longer-running tasks 3

3

. Crypto-native AI agents have also gained traction, with autonomous systems managing wallets, trades, and digital assets. The AI-agent token market reached approximately $22.8 billion, while Virtual Protocol recorded more than 23,500 active wallets and around $479 million in AI-driven economic activity as of March 2026 3

3

. For CFOs and merchants, the x402 protocol offers a glimpse into a new B2B reality where software usage, data calls, model inference, and machine-to-machine work become payable events inside normal business operations 4

4

.