The Pulse
Robinhood Puts AI Trading Agents Inside Its App
Robinhood announced in Houston on September 29 that customers can build AI agents inside its app to research markets, develop strategies and trade. The company says more than 150,000 customers already opened agentic trading accounts, and ag
AI.info Team · Written with AI ·
“We're making Robinhood the best place in the world for active traders by delivering tools once reserved for hedge funds, big banks, and quant firms.”
Vlad Tenev, CEO of Robinhood
Robinhood moves agents into its app
Robinhood announced Robinhood Agents on September 29 at its HOOD Summit in Houston, bringing AI-powered research and automated trading into the brokerage’s app. Customers can build an agent to analyze markets, develop strategies and place trades, without first connecting a separately configured agent through a technical interface. The company also introduced Agent Apps, which add data and specialized tools from 11 third-party providers.
The launch extends Robinhood’s May release of Agentic Trading, which let customers connect external AI agents through the Model Context Protocol, or MCP. That earlier setup required customers to choose an AI platform and connect it to Robinhood. The new in-app product is intended to lower that setup barrier, though Robinhood says access is beginning to roll out and may reach customers over time.
150,000 accounts and nearly 30 million tool calls
Robinhood says more than 150,000 customers have opened agentic trading accounts since the May launch. The company also says agents now use its tools almost 30 million times a day. Those are company-reported figures; the announcement does not provide a breakdown of how many tool calls result in trades, or how the agents’ investment performance compares with human-managed accounts.
The distinction matters: calling a market-data or portfolio tool is not the same as executing an order. Still, the figures show that customers have already connected agents at meaningful scale, before Robinhood placed its own agent-building experience directly in the app. The company’s claim about a much broader potential audience should not be confused with immediate access for every customer: Robinhood says the new features are rolling out in stages.
Separate accounts and approval settings
Robinhood’s safeguards center on where an agent can trade and whether it needs permission first. An agent can trade only in its dedicated account, not in a customer’s other Robinhood accounts. Users can keep trade approvals on, requiring confirmation for orders, or turn them off to allow automatic execution; they can also pause automations or disconnect an agent.
The controls limit an agent’s access, but they do not ensure that its decisions are sound. Robinhood warns that AI agents can make errors, misread instructions or act on incomplete or outdated information, and that customers may lose their entire investment. The company also says it does not control, supervise, monitor, recommend or audit the agents, and customers assume the risks of trades and of sharing account data with third-party AI providers.
Automation without a performance record
The product gives customers a path from asking an agent to research an investment to authorizing it to place an order. That convenience also makes the quality of instructions, model behavior and approval settings part of the investing process. Robinhood’s own disclosures say AI-driven strategies can perform poorly and may be difficult to monitor or stop quickly.
At launch, the clearest measure of adoption is account openings and tool calls—not returns. Robinhood’s new interface puts automated trading within reach of customers who may not have used MCP or built an external agent, while leaving investment decisions and their consequences with the account holder.