Robinhood has announced Robinhood Agents, a built-in artificial intelligence feature that can research markets, analyze portfolios, develop trading strategies and place orders on a customer's behalf. The launch, unveiled at the company's HOOD Summit 2026 conference, moves agentic AI trading from an experimental add-on to a mainstream product aimed at the brokerage's mass-market customer base, and it arrives alongside plans for 24/7 U.S. equity trading, perpetual crypto futures, company-specific earnings contracts and higher intraday margin.
The rollout is the culmination of a bet Robinhood first placed in May 2026, when it began letting technically sophisticated customers connect third-party AI agents to their accounts. According to company figures cited in coverage by Pulse 2.0, more than 150,000 customers have opened agentic trading accounts since then, and those agents are now using Robinhood's tools nearly 30 million times per day. Robinhood Agents is designed to bring that capability to everyone else — no external infrastructure required. For readers tracking how quickly automation is reshaping consumer finance, this is one of the clearest signals yet in our AI industry coverage.
How Robinhood Agents Actually Work
Customers start by naming an agent and opening a dedicated agentic trading account. They then choose an underlying model from several AI providers, including OpenAI; the company said usage of OpenAI's GPT-Luna model will be free through the end of 2026. The agent can only operate with funds held in that dedicated account, a structural separation meant to wall off automated trading from the rest of a customer's portfolio.
Guardrails are part of the default configuration. A trade-approval setting is switched on by default, meaning customers must manually approve every transaction the agent proposes. Users who want full automation can disable the setting, after which the agent can execute orders without per-trade sign-off.
The more ambitious piece is a feature called Loops. A Loop turns a trading strategy into a standing instruction that an agent can execute repeatedly, around the clock. Robinhood's own examples include an agent that checks markets each morning and trades when specified conditions occur, or one that monitors overnight sessions for opportunities. Once activated, Loops can run in the background and may place, modify or cancel trades while the user is away from the device. The company emphasized that customers remain responsible for the automation rules they configure, and that any Loop can be paused or disabled at any time.
'Tools' That Were Once Reserved for Wall Street
Robinhood CEO Vlad Tenev framed the launch as a democratization play, telling Yahoo Finance that the company has given everyday traders tools previously reserved for hedge funds and big banks. The framing is consistent with the company's broader push to court active traders: the HOOD Summit lineup also included plans for true 24/7 U.S. stock trading, perpetual futures on crypto, earnings contracts tied to individual companies, and expanded intraday margin.
The Wall Street Journal, which reviewed the announcement, described the product as "AI bots that trade while you sleep." Fortune noted that the agents are rolling out to millions of users at once, and Startup Fortune reported that the feature covers Robinhood's roughly 29 million users.
The Risk Question Nobody Can Answer Yet
Not everyone is convinced the revolution will be gentle. CoinDesk's coverage carried a blunter assessment: Robinhood has unveiled an AI that trades your money 24/7, "and you carry all the risk." That framing gets at the uncomfortable core of agentic trading. When a model misreads a market, whipsaws in and out of positions, or follows a flawed standing instruction for days before anyone notices, the losses land on the customer, not on Robinhood or the AI provider.
There are also structural concerns that regulators have not yet sorted out. Automated trading systems have existed for decades, but consumer-facing agents that chain together research, strategy and execution raise novel questions about suitability, oversight and what happens when millions of agents simultaneously react to the same market signal. Robinhood's decision to make trade approval the default will dampen some of those worries, but the opt-out is a single toggle away, and the entire product direction points toward hands-off automation.
The market context matters, too. Agentic AI has been the industry's dominant theme this year, with brokerages, banks and payment companies all racing to embed autonomous capabilities in consumer products. Robinhood's version stands out mostly for its scale: 150,000 early adopters became a proof of concept, and the company is now pointing its entire user base at the same on-off switch for automated trading.
What Comes Next
Robinhood says customers can start naming agents and opening agentic accounts now, with model choice expanding beyond OpenAI over time. The 24/7 equity trading, perpetual futures and earnings contracts will follow on the timeline the company laid out at the summit.
The deeper story is what this does to retail investing norms. A generation of traders who learned markets through meme-stock volatility will now share their accounts with models that never sleep, never second-guess themselves and never panic-sell for emotional reasons — or, just as plausibly, panic-sell for statistical ones. Whether Robinhood Agents turns out to be a genuine leveling of the playing field or a machine for generating new categories of retail losses will depend on how those defaults, limits and Loops hold up when markets get rough.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
