Robinhood launches beta support for AI agentic trading and payments

Robinhood has unveiled a product called Agentic Trading, which lets users create a separate, dedicated brokerage account and connect third-party AI agents to it. Those agents can then query portfolio data, place trades, and execute strategies on the user’s behalf, all through a standardized protocol that works with platforms like Claude, ChatGPT, Codex, and Cursor.

How it actually works

The system runs on the Model Context Protocol, an open standard that allows AI platforms to interact with external services in a structured way. Any MCP-compatible service can plug into the endpoint and start operating.

Users set up a separate self-directed account specifically for agent activity. They pre-load it with a balance, connect their preferred AI platform, and define what the agent is allowed to do. The range of automation is flexible, from simple tasks like querying current holdings to more complex operations like automated portfolio rebalancing and price-triggered buy orders.

Agents can also provide risk assessments, but Robinhood is clear that those assessments are strictly informational. Users can review actions before execution if they choose to configure that level of oversight, but the product is designed to function with varying levels of autonomy.

Robinhood frames this as an early-stage offering and has published dedicated support documentation for the feature.

The fine print on responsibility

Users of Agentic Trading hold full responsibility for every trade an agent executes. The liability structure means Robinhood is providing infrastructure, not investment advice, while users retain all the downside risk of those decisions.

Robinhood also offers its own AI assistant called Cortex, which provides research insights and action prompts to Gold subscribers. Cortex is designed to enhance decision-making rather than replace it. The two products together suggest Robinhood is building a layered AI strategy: Cortex for guidance, Agentic Trading for execution.

What this means for investors and the broader market

Strategies like automated rebalancing, conditional order execution, and portfolio monitoring have traditionally required either expensive software or a financial advisor. Now a retail investor can instruct a compatible AI platform to rebalance their portfolio whenever allocation drifts beyond a certain threshold, and the agent handles the rest.

Robinhood already supports crypto trading on its platform. If agents can execute crypto trades with the same autonomy as equity trades, it opens the door to 24/7 automated crypto portfolio management through a regulated brokerage.

Robinhood’s decision to make the account separate from a user’s primary holdings is a meaningful guardrail, sandboxing the AI’s activity so it cannot affect a user’s broader holdings.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.