Robinhood is giving customers access to AI agents that can independently research markets, create trading strategies and execute orders, potentially around the clock. The company, however, says customers remain responsible for the risks and losses associated with trades carried out by the technology.
The Nasdaq-listed brokerage unveiled “Robinhood Agents” Tuesday at its HOOD Summit in Houston. Built directly into the Robinhood app, the feature allows users to choose an AI agent and define the permissions and limits under which it can trade.
Robinhood has more than 27 million funded accounts. While a traditional chatbot primarily provides answers, an AI agent can take actions for the user. Robinhood’s agents can therefore analyze markets and buy or sell assets on a customer’s behalf based on the instructions and controls selected by that customer.
The rollout extends automated trading capabilities that have historically been used by hedge funds and quantitative trading firms to individual investors.
Robinhood previously introduced agentic trading in May, allowing more technically advanced customers to connect their own AI agents to their accounts. Since then, more than 150,000 customers have opened agentic trading accounts, while those agents have been using Robinhood’s tools nearly 30 million times a day, according to the company.
The latest offering allows customers without the same technical expertise to select and authorize an AI agent directly.
“Agentic accounts come with trade approvals settings which you can configure to allow automated trade execution. With approvals on, your agent cannot place an order until you approve it. You can turn trade approvals off, and if you do, your agent can place orders without asking you to confirm each one,” Robinhood said.
Robinhood is also developing a feature called Loops that will allow users to turn a strategy into an ongoing instruction. The agent could repeatedly check market conditions and execute trades when predefined criteria are met.
For instance, a user could set an agent to review the market each morning and trade when specific conditions occur, or leave the strategy running overnight. Robinhood said Loops will be available soon.
The move comes as AI agents are increasingly being designed to perform financial tasks without direct human intervention. Meta’s Muse assistant can access users’ bank balances and investments, while Coinbase-developed x402 enables agents to pay for services using stablecoins.
Customers bear the trading risk
Robinhood’s disclosures place responsibility for AI-driven trading firmly on the customer.
Users “assume all risk for trades executed by AI agents and for any use of your data by third-party LLM providers,” according to the company. Robinhood also said it “does not control, supervise, monitor, recommend, or audit agents.”
The company has issued additional warnings about Loops because the feature can continue executing a strategy without requiring individual approval for every order.
Once enabled, Loops “may place, modify, or cancel trades in your account automatically, without prompting you for approval on each transaction – including while you’re asleep, away from your device, or otherwise not monitoring the market.”
The feature will follow a customer’s instructions “exactly as configured, including during periods of market volatility.”
Robinhood does not guarantee how Loops will perform under any particular market conditions and says automated trading carries the same risks as manually placing trades. Users can deactivate the feature, but transactions that have already been executed by Loops will not automatically be reversed.
Autonomous trading raises wider concerns
The risks identified by Robinhood are focused mainly on individual customers, while regulators and researchers have raised broader questions about what could happen if large numbers of AI agents begin trading simultaneously.
Bank of England Deputy Governor Sarah Breeden said in June that autonomous AI agents could “amplify volatility in stress” and potentially lead to a “market meltdown.” She also warned that existing financial regulation was not designed for agentic systems.
One potential issue is coordinated market behavior. If numerous agents interpret the same news similarly and place trades at roughly the same time, their combined activity could magnify a market move.
A study by researchers from Wharton and the Hong Kong University of Science and Technology found that AI trading agents in a simulated environment were capable of colluding and fixing prices to generate collective profits, even without an explicit communication channel.
The researchers also found that agents could sustain above-market profits without communication, agreement or intent, highlighting potential difficulties for regulators overseeing autonomous systems.
These risks remain largely theoretical for now, given that agentic trading is still an emerging technology with relatively limited adoption. The regulator warnings and academic findings apply to AI-powered trading generally and do not show that Robinhood’s agents display the same behavior.
Still, the research underscores that automated trading can introduce risks beyond those associated with simply giving investors a more convenient way to execute trades.
