Robinhood expanded its Agentic Trading platform to support cryptocurrency markets as of early July 2026, allowing third-party AI agents to buy and sell digital assets 24/7 on behalf of users. The feature originally launched on May 27, 2026, but was limited to equities. Now it covers the asset class that literally never sleeps.

How it works

The system lets users create dedicated agentic accounts where AI agents operate under strict, user-defined parameters. These AI agents connect to Robinhood’s Trading MCP, which stands for Model Context Protocol. That’s the infrastructure layer that feeds real-time market data to the agents and gives them the ability to execute trades.

Built-in safety features require user authorization for agent actions, and all activity remains visible to the user. The 24/7 capability is where this gets genuinely interesting for crypto specifically. Stock markets close at 4 PM Eastern and take weekends off. Crypto does not. An AI agent that monitors and trades around the clock addresses that gap directly.

Vlad Tenev’s AI bet

Robinhood CEO Vlad Tenev has stated that AI agents will soon replicate every capability a human trader has, and potentially exceed them.

The crypto expansion is part of a larger strategic push Robinhood announced in early July 2026. Alongside AI trading, the company outlined plans for tokenized investing, blockchain initiatives under the Robinhood Chain banner, and decentralized finance products. Notably, no specific tokens are linked to these announcements.

What this means for crypto traders and the broader market

For retail investors, strategies that previously required either constant screen time or expensive third-party bot subscriptions could now be accessible through a familiar interface. Automated trading has historically been the domain of hedge funds and prop desks with engineering teams. Robinhood is packaging that capability for someone with a few hundred dollars and a phone.

AI agents operating in volatile crypto markets with 24/7 execution capability can amplify losses just as effectively as they can capture gains. User-defined limits provide guardrails, but a poorly configured AI agent in a flash crash scenario could execute trades that a human would instinctively avoid.