Coinbase launched “Coinbase for Agents” on Wednesday, a new platform that enables artificial intelligence assistants including ChatGPT and Anthropic’s Claude to autonomously trade cryptocurrencies and execute financial transactions within user accounts. The product represents Coinbase’s direct entry into what the company calls “agentic commerce,” where AI systems increasingly handle financial activity on behalf of users.
Users can now authorize AI agents to trade spot cryptocurrencies and derivatives through natural language commands, allowing agents to rebalance portfolios and execute automated trading strategies without human intervention for each transaction. The platform integrates access to Coinbase’s advanced trading tools alongside market data and analytics. Support for equities trading and prediction markets is planned for future releases.
Coinbase embedded its x402 payments protocol into the agent platform, enabling machine-to-machine payments for services including premium research, data APIs, and computing resources. This allows agents to make small payments without subscription requirements or manual checkout processes. The architecture removes friction from agent-executed transactions while maintaining security guardrails.
Addressing security concerns, Coinbase allows agents to operate within isolated portfolios with customizable spending caps, trade limits, and restrictions on which services agents can access. These controls ensure users maintain authority over their accounts even when AI systems execute transactions autonomously. The approach balances automation convenience with user protection.
Coinbase cited analyst forecasts suggesting autonomous agents could account for up to 20 percent of e-commerce activity by 2030, positioning the company to capture significant market share in AI-driven financial commerce. The launch follows Robinhood’s recent announcement of an AI agent trading product, showing the competitive pressure driving fintech platforms toward autonomous financial services.
Industry observers note that autonomous agent accounts represent a logical evolution of AI integration into financial markets. The ability for language models like Claude and ChatGPT to access real-time market data and execute trades expands their practical utility beyond analysis and recommendation into direct market participation.
