Crypto’s next growth wave, Coinbase argues, will come from AI agents transacting with each other in an agent-to-agent economy, with Base leading activity and over 100 million agent payments anticipated in early 2026 according to Chainalysis.

Coinbase promotes agentic payments as central to this growth, embedding agent interactions within the Base chain and helping set standards to enable automated, machine-to-machine commerce.

The thesis positions autonomous AI agents as the primary drivers of crypto expansion, shifting emphasis from human-led trades to programmable economic agents.

Users should sandbox exposure by funding a separate portfolio and granting agents limited permissions to reduce risk from unintended orders.

A robust infrastructure model is needed—covering value, payments, identity, markets, and tight, accountable permissions—while addressing risks like unauthorized transactions, AI fraud, deepfakes, market manipulation, privacy loss, identity disputes, and harm responsibility.

Current platform enables spot trading across 900+ crypto pairs, with eligible US users able to access futures, equities, portfolio monitoring, and USDC-dollar conversions under regulatory and account constraints.

Asset tokenization (stocks, bonds, commodities, property, IP) is viewed as a way to represent markets for AI agents to analyze, exchange, rebalance, and manage digital assets.

Bitcoin is framed as scarce neutral money and a potential settlement layer, while stablecoins provide continuous rails and programmable payments for work-based or condition-based transfers.

Investors are advised to target Ethereum and Solana for exposure to agentic growth due to their broader ecosystems, rather than niche agent-specific tokens, offering diversification and downside protection.

Coinbase for Agents currently enables AI applications to access designated portfolios through Coinbase Advanced Trade, via a remote server or local CLI, under a sub-account with user-defined transaction limits.

Recent developments include SEC registrations for US stock perpetuals and leadership changes, with ongoing questions about identity checks, disputes, refunds, and legal responsibility when software initiates transactions.

Regulatory considerations are rising around autonomous financial interactions, including digital identity, AML standards, and oversight as code-driven agents transact.